阅读说明：本技术 基于机器视觉的水体溶解氧检测装置 (Water body dissolved oxygen detection device based on machine vision ) 是由 饶宾期 钱浩 董文杰 苏小雨 逄鸿儒 于 2019-10-29 设计创作，主要内容包括：本发明公开了一种基于机器视觉的水体溶解氧检测装置,包括：取样装置,从待测水体中吸取水样,并存储于储样瓶中；滴液装置,存储有多种滴定的溶液或试剂,用于与水样中的溶解氧反应；石英玻璃反应皿,用于接收所述储样瓶中输出的水样和滴液装置中输出的溶液或试剂,并在石英玻璃反应皿中进行反应；信息处理模块,包括拍摄模块和控制面板,通过所述拍摄模块对所述石英玻璃反应皿内液体进行拍照,拍照获得的图像输入至所述控制面板的处理器,由所述处理器进行处理,提取图像的RGB分量,并将RGB分量转换成HSI分量后,之后将HSI分量作为输入量输入训练好的神经网络中得出溶解氧浓度；所述神经网络通过误差反向传播方法进行训练而成。(The invention discloses a machine vision-based water body dissolved oxygen detection device, which comprises: the sampling device is used for sucking a water sample from the water body to be detected and storing the water sample in the sample storage bottle; the dropping device stores various titrated solutions or reagents and is used for reacting with dissolved oxygen in the water sample; the quartz glass reaction vessel is used for receiving the water sample output from the sample storage bottle and the solution or reagent output from the dropping device and carrying out reaction in the quartz glass reaction vessel; the information processing module comprises a shooting module and a control panel, the liquid in the quartz glass reaction vessel is shot through the shooting module, an image obtained through shooting is input to a processor of the control panel, the image is processed by the processor, RGB components of the image are extracted, the RGB components are converted into HSI components, and then the HSI components are input into a trained neural network as input quantities to obtain the concentration of dissolved oxygen; the neural network is trained by an error back propagation method.)

1. A machine vision-based water body dissolved oxygen detection device comprises:

the sampling device is used for sucking a water sample from the water body to be detected and storing the water sample in the sample storage bottle;

the dropping device stores various titrated solutions or reagents and is used for reacting with dissolved oxygen in the water sample;

the quartz glass reaction vessel (15) is used for receiving the water sample output from the sample storage bottle and the solution or reagent output from the dropping device, and reacting in the quartz glass reaction vessel (15);

the information processing module comprises a shooting module and a control panel, liquid in the quartz glass reaction vessel (15) is shot through the shooting module, an image obtained through shooting is input to a processor of the control panel, the image is processed by the processor, RGB components of the image are extracted, the RGB components are converted into HSI components, and then the HSI components are input into a trained neural network as input quantities to obtain the concentration of dissolved oxygen; the neural network is trained by an error back propagation method.

2. The machine vision-based water body dissolved oxygen detection device according to claim 1, wherein: the sampling device comprises a water suction pipe rotating disc (22-6), a sample storage bottle (2), a water suction pump (21), a weight increasing block (24) and a water suction pipe (25); the weight increasing block (24) is arranged on the water suction pipe (24) so that buoyancy can be counteracted to enter the river channel; the water suction pipe rotating disc (22-6) is connected with the weight increasing block (24) through the water suction pipe (24), the water suction pipe (24) is driven to a specified depth through the gravity of the weight increasing block (25) according to a certain number of turns of depth required by setting, and then the water suction pump (21) works to pump water into the sample storage bottle (2) until the water inlet of the sample storage bottle (2) continuously overflows water for two minutes.

3. The machine vision-based water body dissolved oxygen detection device according to claim 2, wherein: the water suction pipe (25) and a liquid pipeline communicated with the water suction pipe are arranged on a plurality of sections of pull rods, the pull rods and the contraction rods of the plurality of sections of pull rods are realized by the rod moving device, the rod moving device comprises a sampling motor (17), a rolling screw rod (18), a sampling sliding block (19), a sampling guide rail (20) and a gas claw (23), and a nut is fixed on the sampling sliding block (19); a sampling motor (17) drives a rolling screw rod (18) to control a sampling slide block (19) to move back and forth on a sampling guide rail (20) so as to move the position of the gas claw; the air claw (23) is connected with the sampling sliding block (19), the air claw (23) comprises an air claw arm and an air claw, the air claw is inflated in the initial state, the two arms are opened and move in the same position with the sampling sliding block (19), after the air claw reaches a designated position, the two arms of the claw are closed downwards, and after the air claw is closed, the air claw is vacuumized, so that the air claw is tightly attached to the outer wall of the multi-section pull rod (22) and the rod moving is realized.

4. The machine vision-based water body dissolved oxygen detection device according to claim 3, wherein: the liquid dropping device comprises a displacement device, the displacement device comprises a stepping motor (6), a turntable (7), a liquid dropping guide rail (10) and a liquid dropping slider (33), the turntable (7) is fixedly connected with the liquid dropping guide rail (10), so that the liquid dropping guide rail (10) can rotate along with the turntable, the stepping motor (6) controls the turntable (7) to rotate, and meanwhile, the liquid dropping guide rail is driven to rotate by a preset angle; four total guide rails of dropping liquid guide rail (10), fix respectively on four faces of dropping liquid guide rail (10) center pin, dropping liquid slider (33) including sliding part and fixed part, sliding part makes it slide on dropping liquid guide rail (10), fixed part is used for fixed dropping liquid mechanism, makes dropping liquid mechanism can remove along with dropping liquid slider (33), four dropping liquid mechanisms are fixed respectively to four dropping liquid sliders (33), makes dropping liquid mechanism arrive the assigned position through displacement device.

5. The machine vision-based water body dissolved oxygen detection device according to claim 4, wherein: dropping liquid mechanism include liquid storage pot (8), dropping liquid solenoid valve (9) and dropping liquid pipe, dropping liquid mechanism a total 4, every dropping liquid mechanism all has different reagent, the reagent is stored in the liquid storage pot (8) of dropping liquid mechanism, the volume in quartz glass reaction ware (15) is instiled into to different solution or reagent through dropping liquid solenoid valve (9) control in every dropping liquid mechanism.

6. The machine vision-based water body dissolved oxygen detection device according to claim 5, wherein: the quartz glass reaction vessel (15) comprises a liquid dropping port (15-1), a sample conveying sealing port (15-2), a liquid discharging port (15-3), an air pumping port (15-4) and a liquid transferring port (15-5), the quartz glass reaction vessel (15) is made of acid and alkali resistant quartz glass, the liquid dropping port (15-1), the sample conveying sealing port (15-2), the liquid discharging port (15-3) and the air pumping port (15-4) are provided with an electromagnetic valve to play a role of opening and closing, and the liquid dropping port (15-1), the sample conveying sealing port (15-2), the liquid discharging port (15-3) and the air pumping port (15-4) are in a closed state in an initial state; the dropping liquid port (15-1) is used for dropping a solution or a reagent into the reaction vessel by a dropping device, the surface of the sample conveying sealing port (15-2) is provided with a layer of sealing ring which is combined with the sample conveying port (11-3) to play a role of sealing and isolating external air, a water sample is transferred into the quartz glass reaction vessel from the sample storage bottle (2) through the sample conveying sealing port (15-2) by adopting a transfusion device, and the air in the reaction vessel is firstly pumped out before chemical operation so as to keep the vacuum state in the reaction vessel.

7. The machine vision-based water body dissolved oxygen detection device according to claim 6, wherein: the sampling device conveys water samples into the quartz glass reaction vessel (15) through the infusion part, the infusion part comprises a water outlet electromagnetic valve (4), a liquid exchanger (5) and a sample conveying device (11), the water outlet electromagnetic valve (4) has two liquid-infusion quantities which respectively control a sample storage bottle (2) and a water storage bottle (3) into the quartz glass reaction vessel, the water storage bottle (3) is used for containing deionized water which plays a role of dilution, the liquid exchanger (5) enables the sample storage bottle (2) and the water storage bottle (3) to be infused into the quartz glass reaction vessel through the same set of sample conveying device (11), the sample conveying device (11) comprises a sample conveying pipe shell (11-1), an extension pipe (11-2), a sample conveying port (11-3) and a sample conveying pipe (11-4), the tail part of the sample conveying pipe (11-4) is connected with the liquid exchanger (5), the sample conveying pipe shell (11-1) is used for ensuring that the sample conveying device (11) is perpendicular to the reaction vessel infusion sealing port (15-2), the extension pipe (11-2) extends towards the gravity direction in the infusion working state to enable the sample conveying port (11-3) to be in sealing contact with the infusion sealing port (15-2), and the sample conveying port (11-3) discharges the sample conveying pipe (11-4) after reaching a designated position to enable the sample conveying pipe to be in contact with the bottom of the quartz glass reaction vessel.

8. The machine vision-based water body dissolved oxygen detection device according to claim 7, wherein: the liquid transfer port (15-5) transfers a certain amount of solution to a spectrophotometer (34) through a liquid transfer electromagnetic valve (35); the spectrophotometer (34) is used for reading the iodine concentration in the iodine solution after the chemical reaction and taking the iodine concentration as one of the input items of the neural network; quantitatively acquiring a solution in a reaction vessel through a liquid-transferring electromagnetic valve (35), and injecting the acquired solution into a quartz glass cuvette in a sample grid in a spectrophotometer (34) through a liquid transferring pipe; the obtained iodine amount concentration information is transmitted to the control panel (1) as an input item of the neural network.

9. The machine vision-based water body dissolved oxygen detection device according to claim 7, wherein: the shooting module comprises a darkroom, a voltage-stabilizing light source (14) and a camera (16), wherein the darkroom consists of a shading tail plate (28), a shading device (29) and the rest of wall surfaces in one layer of the cabinet body coated with black light-absorbing pigment, and the shading device (29) is divided into an upper plate and a lower plate to form a closed darkroom space; the camera (16), the voltage-stabilizing light source (14) and the quartz glass reaction vessel (15) are positioned at the same height in the center in the shooting state; the voltage-stabilizing light source (14) is a long-surface screen type light source, and the brightness of the voltage-stabilizing light source is adjusted through the control panel (1); the camera (16) shoots the quartz glass reaction vessel (15) in the process of dissolving and precipitating through chemical operation, transmits the obtained picture to a processor of the control panel (1), and judges whether the quartz glass reaction vessel is completely dissolved or not through image processing; after the reaction is completed, the camera (16) shoots the quartz glass reaction vessel (15) in a dark room, and the obtained picture is transmitted to a processor of the control panel (1) for image processing.

10. The machine vision-based water body dissolved oxygen detection device according to claim 7, wherein: the dropping liquid device firstly drops into manganese sulfate solution in quartz glass reaction vessel (15), and the sodium hydroxide-potassium iodide solution that again drops into produces the sediment after both reactions, and the concentrated sulfuric acid solution that again drops into is in order to dissolve the sediment and to change out the iodine, and the concentration of iodine has characterized the concentration of dissolved oxygen, makes the mixed solution discolour through adding starch solution, handles the image through shooing to through neural network processing acquisition dissolved oxygen concentration.

Technical Field

The invention relates to the field of water quality detection, in particular to a water body dissolved oxygen detection device based on machine vision.

Background

Disclosure of Invention

In view of the above, the present invention provides a dissolved oxygen detection device capable of implementing water dissolved oxygen detection based on machine vision and neural network, and capable of automatically sampling and automatically performing chemical operation, so as to solve the problems in the prior art.

The technical scheme of the invention is to provide a machine vision-based water body dissolved oxygen detection device with the following structure, which comprises:

the sampling device is used for sucking a water sample from the water body to be detected and storing the water sample in the sample storage bottle;

the dropping device stores various titrated solutions or reagents and is used for reacting with dissolved oxygen in the water sample;

the quartz glass reaction vessel is used for receiving the water sample output from the sample storage bottle and the solution or reagent output from the dropping device and carrying out reaction in the quartz glass reaction vessel;

the information processing module comprises a shooting module and a control panel, the liquid in the quartz glass reaction vessel is shot through the shooting module, an image obtained through shooting is input to a processor of the control panel, the image is processed by the processor, RGB components of the image are extracted, the RGB components are converted into HSI components, and then the HSI components are input into a trained neural network as input quantities to obtain the concentration of dissolved oxygen; the neural network is trained by an error back propagation method.

Optionally, the sampling device comprises a water suction pipe rotating disc, a sample storage bottle, a water suction pump, a weight increasing block and a water suction pipe; the weight increasing block is arranged on the water suction pipe, so that buoyancy can be counteracted to enter the interior of the river channel; the rotating disc of the water suction pipe is connected with the weight increasing block through the water suction pipe, the rotating disc rotates for a certain number of turns according to the set required depth, the water suction pipe is pulled to the specified depth through the gravity of the weight increasing block, and then the water suction pump works to pump water into the sample storage bottle until the water continuously overflows from the water inlet of the sample storage bottle for two minutes.

Optionally, the water suction pipe and the liquid pipeline communicated with the water suction pipe are arranged on a plurality of sections of pull rods, the pull rods and the contraction rods of the plurality of sections of pull rods are realized by the rod moving device, the rod moving device comprises a sampling motor, a rolling screw rod, a sampling slide block, a sampling guide rail and a gas claw, and a nut is fixed on the sampling slide block; the sampling motor drives the rolling screw rod to control the sampling slide block to move back and forth on the sampling guide rail so as to move the position of the air claw; the air claw is connected with the sampling sliding block and comprises an air claw arm and an air gripper, the air gripper is inflated in an initial state, the two arms are opened and move in the same position with the sampling sliding block, after the air gripper reaches a designated position, the two arms of the gripper are closed downwards, and after the air gripper is closed, the air gripper is vacuumized, is tightly attached to the outer wall of the multi-section pull rod, and moves the rod.

Optionally, the liquid dropping device comprises a displacement device, the displacement device comprises a stepping motor, a turntable, a liquid dropping guide rail and a liquid dropping slider, the turntable is fixedly connected with the liquid dropping guide rail, so that the liquid dropping guide rail can rotate along with the turntable, and the stepping motor controls the turntable to rotate and simultaneously drives the liquid dropping guide rail to rotate by a preset angle; four total guide rails of dropping liquid guide rail, fix respectively on four faces of dropping liquid guide rail center pin, the dropping liquid slider include sliding part and fixed part, sliding part makes it slide on the dropping liquid guide rail, fixed part is used for fixed dropping liquid mechanism, makes dropping liquid mechanism can remove along with the dropping liquid slider, four dropping liquid mechanisms are fixed respectively to four dropping liquid sliders, makes dropping liquid mechanism arrive the assigned position through displacement device.

As optional, dropping liquid mechanism include liquid storage pot, dropping liquid solenoid valve and dropping liquid pipe, dropping liquid mechanism a total of 4, every dropping liquid mechanism all has different reagent, the reagent is stored in the liquid storage pot of dropping liquid mechanism, the volume in the quartz glass reaction vessel is dripped into to different solution or reagent of dropping liquid solenoid valve control on through every dropping liquid mechanism.

Optionally, the quartz glass reaction vessel comprises a liquid dropping port, a sample conveying sealing port, a liquid discharging port, an air pumping port and a liquid transferring port, the quartz glass reaction vessel is made of acid-alkali-resistant quartz glass, the liquid dropping port, the sample conveying sealing port, the liquid discharging port and the air pumping port are all provided with an electromagnetic valve which plays a role of opening and closing the ports, and the liquid dropping port, the sample conveying sealing port, the liquid discharging port and the air pumping port are all in a closed state in an initial state; the dropping liquid mouth be arranged in the dropping liquid device with solution or reagent instils into the reaction ware, defeated appearance sealing port surface have one deck sealing washer, combine to play sealed isolated outside air's effect with defeated appearance mouth, adopt infusion set from storage appearance bottle and through defeated appearance sealing port with the water sample from storage appearance bottle transfer to quartz glass reaction ware in, the extraction opening connect an aspiration pump, at first take out the inside air of reaction ware before chemical operation, make and keep vacuum state in the reaction ware.

Optionally, the sampling device conveys a water sample into the quartz glass reaction vessel through the liquid conveying part, the liquid conveying part comprises a water outlet electromagnetic valve, a liquid exchanger and a sample conveying device, the water outlet electromagnetic valve is provided with two liquid conveying amounts which respectively control a sample storage bottle and a water storage bottle into the quartz glass reaction vessel, the water storage bottle is used for containing deionized water which plays a role in dilution, the liquid exchanger enables the sample storage bottle and the water storage bottle to convey the liquid into the quartz glass reaction vessel through the same set of sample conveying device, the sample conveying device comprises a sample conveying pipe shell, a telescopic pipe, a sample conveying port and a sample conveying pipe, the tail part of the sample conveying pipe is connected with the liquid exchanger, the sample conveying pipe shell is used for ensuring that the sample conveying device is vertical to the liquid conveying sealing port of the reaction vessel, the telescopic pipe stretches out towards the gravity direction in the working state of the liquid conveying, the sample conveying port is in sealing contact with the liquid conveying sealing port, and the sample conveying port can convey the sample pipe after reaching a designated position, so that the quartz glass reaction vessel is contacted with the bottom of the quartz glass reaction vessel.

Optionally, the pipetting port delivers a certain amount of solution to the spectrophotometer through the pipetting electromagnetic valve; the spectrophotometer is used for reading the iodine concentration in the iodine solution after the chemical reaction and taking the iodine concentration as one of the input items of the neural network; quantitatively obtaining the solution in the reaction vessel through a liquid transfer electromagnetic valve, and injecting the obtained solution into a quartz glass cuvette in a sample grid in a spectrophotometer through a pipette; and transmitting the obtained iodine amount concentration information to a control panel as an input item of the neural network.

Optionally, the shooting module comprises a darkroom, a voltage-stabilizing light source and a camera, the darkroom comprises a light-shielding tail plate, a light-shielding device and the rest of wall surfaces inside one layer of the cabinet body coated with black light-absorbing pigment, and the light-shielding device is divided into an upper plate and a lower plate to form a closed darkroom space; the center of the camera, the stabilized voltage light source and the quartz glass reaction vessel are positioned at the same height in the shooting state; the voltage-stabilizing light source is a long-surface screen type light source, and the brightness of the voltage-stabilizing light source is adjusted through the control panel; the camera shoots the quartz glass reaction vessel in the process of dissolving and precipitating through chemical operation, transmits the obtained picture to a processor of a control panel, and judges whether the quartz glass reaction vessel is completely dissolved or not through image processing; after the reaction is completed, the camera shoots the quartz glass reaction vessel in a dark room, and transmits the obtained picture to a processor of a control panel for image processing.

Optionally, the dropping device firstly drops a manganese sulfate solution into the quartz glass reaction vessel, then drops a sodium hydroxide-potassium iodide solution, the manganese sulfate solution and the sodium hydroxide-potassium iodide solution react to generate a precipitate, and then drops a concentrated sulfuric acid solution to dissolve the precipitate and replace iodine, wherein the concentration of the iodine represents the concentration of dissolved oxygen, the mixed solution is discolored by adding a starch solution, an image is processed by photographing, and the concentration of the dissolved oxygen is obtained by processing through a neural network.

By adopting the structure, compared with the prior art, the invention has the following advantages: (1) the automatic titration device can automatically realize part of the operation process of the iodometry, so that the process is more precise, improper manual operation is avoided, and the danger of using chemicals manually is reduced; (2) the adopted river channel water sampling device can adjust the telescopic length of the river channel water sampling device according to the width of different river channels so as to collect water samples at different positions of the river channel, and meanwhile, the length of the water suction pipe can be adjusted according to different requirements on depth so as to collect water samples at different depths of the river channel; (3) the machine vision of the sampling can analyze whether the precipitate is completely dissolved after a certain time after the quantitative sulfuric acid is added through image processing after the picture is transmitted into the control panel, and if the precipitate is not dissolved, the control panel sends a signal to the titration device to continuously add a certain amount of sulfuric acid; (4) the soft measurement modeling method for the water quality of the sewage based on the machine vision and the neural network replaces the titration of sodium thiosulfate in an iodometry method, directly shoots a sample solution added with a starch solution after dilution, reduces the use of an easily-exploded hazardous chemical potassium dichromate, reduces the use of the sodium thiosulfate and saves the time at the same time.

Drawings

FIG. 1 is a front view of a machine vision-based water body dissolved oxygen detection device according to the present invention;

FIG. 2 is an isometric view of a machine vision-based water body dissolved oxygen detection device of the present invention;

FIG. 3 is a rear view of the water body dissolved oxygen detection device based on machine vision according to the present invention;

FIG. 4 is a three-dimensional view of a sampling device of the machine vision-based water body dissolved oxygen detection device of the invention;

FIG. 5 is a three-dimensional view of a water suction pipe turntable of the machine vision-based water body dissolved oxygen detection device;

FIG. 6 is a three-dimensional view of a storage battery of the water body dissolved oxygen detection device based on machine vision;

FIG. 7 is a schematic structural diagram of a sample transportation device of the water body dissolved oxygen detection device based on machine vision according to the present invention;

FIG. 8 is a three-dimensional view of a quartz glass reaction vessel of the water body dissolved oxygen detection device based on machine vision according to the invention;

FIG. 9 is a three-dimensional view of a liquid dropping device of the water body dissolved oxygen detection device based on machine vision according to the invention;

FIG. 10 is a schematic diagram of a neural network.

As shown in the figure, 1, a control panel, 2, a sample storage bottle, 3, a water storage bottle, 4, a water outlet electromagnetic valve, 5, a liquid changer, 6, a stepping motor, 8, a liquid storage tank, 9, a dropping electromagnetic valve, 10, a dropping guide rail, 11, a sample storage device, 12, a shaking wheel disc, 13, a shaking support, 14, a pressure stabilizing light source, 15, a quartz glass reaction vessel, 16, a camera, 17, a sampling motor, 18, a rolling screw rod, 19, a sampling slide block, 20, a sampling guide rail, 21, a water suction pump, 22, a multi-section pull rod, 23, a gas claw, 24, a weighting block, 25, a water suction pipe, 26, a solar light panel, 27, a solar energy storage electromagnet, 28, a shading tail plate, 29, a shading device, 30, a waste liquid collection box, 31, an air suction pump, 32, a sampling support, 33, a dropping slide block, 34, a spectrophotometer, 35.

Detailed Description

The invention is further described with reference to the following figures and specific examples.

The invention is intended to cover alternatives, modifications, equivalents, and alternatives that may be included within the spirit and scope of the invention. In the following description of the preferred embodiments of the present invention, specific details are set forth in order to provide a thorough understanding of the present invention, and it will be apparent to those skilled in the art that the present invention may be practiced without these specific details. Moreover, the drawings of the present invention are not necessarily to scale, nor are they necessarily to scale, as may be shown and described herein.

As shown in FIG. 1, the detection device for dissolved oxygen in water based on machine vision of the present invention is illustrated, and comprises: control panel 1, storage bottle 2, water storage bottle 3, play water solenoid valve 4, liquid changing device 5, step motor 6, liquid storage pot 8, dropping liquid solenoid valve 9, dropping liquid guide rail 10, storage device 11, rock rim plate 12, rock support 13, steady voltage light source 14, quartz glass reaction vessel 15, camera 16, sampling motor 17, ball screw 18, sampling slider 19, sampling guide rail 20, suction pump 21, multisection pull rod 22, gas claw 23, weight increasing piece 24, water absorption pipe 25, solar light panel 26, solar energy holds electromagnetism 27, shading tailboard 28, shade 29, waste liquid collecting box 30, aspiration pump 31, sampling support 32, dropping liquid slider 33, spectrophotometer 34, move liquid solenoid valve 35, cabinet 36, the sampling part constitute by sampling support 32, multisection pull rod 22, move the pole device and absorb water the device. The multi-section pull rod 22 is fixed on the sampling support 32 and is divided into five sections including a first section rod 22-1, a second section rod 22-2, a third section rod 22-3, a fourth section rod 22-4 and a fifth section rod 22-5, the tail ends of the first section rod to the fourth section rod are provided with outer grooves, the front ends of the second section rod to the fifth section rod are provided with inner grooves, so that dislocation of two adjacent rods during maximum displacement is prevented, the number of sections of the telescopic rods and the telescopic length of the rod with the displacement at the tail end can be adjusted by a rod moving device according to the width of a river channel, and water body samples of all parts of the river channel are convenient to take.

The rod moving device consists of a sampling motor 17, a rolling screw rod 18, a sampling slide block 19, a sampling guide rail 20 and an air claw 23. A nut is fixed on the sampling slide block 19, and the sampling slide block 19 can be controlled to move back and forth on a sampling guide rail 20 by driving a rolling screw rod 18 through a sampling motor 17 so as to move the position of the gas claw; the air claw 23 comprises an air claw arm and a hand grip, the air hand grip is inflated in the initial state, the two arms are opened and move in the same position with the sampling slide block 19, after the air hand grip reaches the designated position, the two arms of the hand grip are closed downwards, and after the air hand grip is closed, the air hand grip is vacuumized, so that the air hand grip can be tightly attached to the outer wall of the rod of the multi-section pull rod 22, and the function of stretching each section of rod can be achieved.

The moving rod action is divided into a pull rod part and a contraction rod part, and the pull rod part comprises: firstly, the gas claw 23 is moved to the front end position of a first section rod 22-1 in the initial stage through the rotation of the sampling motor 17, after the gas claw 23 fixes the first section rod 22-1, the first section rod 22-1 is pulled outwards to a specified position through the driving displacement of the sampling motor 17, then the gas claw 23 is released from the working state and is driven to move reversely to the front end of a second section rod 22-2 through the sampling motor 17, the operation is the same as that for moving the first section rod 22-1 until a fourth section rod 22-4 is completely pulled out of a fifth section rod 22-5, the maximum sampling distance is obtained, and the required sampling distance can be reached through setting the number of stretching section rods and the length of a tail section rod; the rod retracting part is as follows: assuming that the gas claw 23 is at the most distal distance, and is at the most distal position of the fourth rod 22-4, it is pulled back into the fifth rod 22-5 by the displacement of the sampling motor 17, and then the gas claw 23 is released from the working position and is displaced to the front end of the third rod 22-3 by the displacement of the sampling motor 17, which is the same operation as the movement of the fourth rod 22-4, until the first rod 22-1 is completely retracted.

The water absorption device comprises a water absorption pipe rotating disc 22-6, a sample storage bottle 2, a water suction pump 21, a weight increasing block 24 and a water absorption pipe 25. The water suction pipe 25 is made of a material with good toughness and corrosion resistance, is light in weight and is filled with air to float on the water surface, so that measures are needed to help the water suction pipe to enter the river channel; the weight increasing block 24 adds weight to the suction pipe 24, so that buoyancy can be counteracted to enter the river channel; the rotating disc 22-6 of the suction pipe rotates for a certain number of turns according to the set required depth, so that the suction pipe 24 can be pulled to the designated depth through the gravity of the weighting block 25, and then the water suction pump 21 works to pump water into the sample storage bottle 2 until the water is continuously overflowed from the water inlet of the sample storage bottle 2 for two minutes.

The liquid dropping device comprises a displacement device and a liquid dropping mechanism, the displacement device comprises a stepping motor 6, a turntable 7, a liquid dropping guide rail 10 and a liquid dropping slider 33, the turntable 7 is fixedly connected with the liquid dropping guide rail 10, so that the liquid dropping guide rail 10 can rotate along with the turntable, the stepping motor 6 controls the turntable 7 to rotate, and meanwhile, the liquid dropping guide rail is driven to rotate by a preset angle; four total guide rails of dropping liquid guide rail 10, fix respectively on four faces of dropping liquid guide rail 10 center pin, dropping liquid slider 33 include sliding part and fixed part, sliding part makes it slide on dropping liquid guide rail 10, fixed part is used for fixed dropping liquid device, makes the dropping liquid device can remove along with dropping liquid slider 33, four dropping liquid devices are fixed respectively to four dropping liquid sliders 33, can help the dropping liquid device to arrive the assigned position through displacement device.

The dropping liquid mechanism constitute by liquid storage pot 8, dropping liquid solenoid valve 9 and dropping liquid pipe, dropping liquid device a total 4, every dropping liquid device has all deposited different reagent, the reagent is stored in dropping liquid device's liquid storage pot 8, need instil into the volume in the quartz glass reaction vessel 15 through the different solution of dropping liquid solenoid valve 9 control on every dropping liquid device.

The quartz glass reaction vessel 15 comprises a liquid dropping port 15-1, a sample conveying sealing port 15-2, a liquid discharging port 15-3, an air pumping port 15-4 and a liquid transferring port 15-5, the quartz glass reaction vessel 15 is made of acid and alkali resistant quartz glass, the liquid dropping port 15-1, the sample conveying sealing port 15-2, the liquid discharging port 15-3 and the air pumping port 15-4 are all provided with an electromagnetic valve to play a role of opening and closing the ports, and the liquid dropping port 15-1, the sample conveying sealing port 15-2, the liquid discharging port 15-3 and the air pumping port 15-4 are all in a closed state in an initial state; the dropping liquid port 15-1 is used for the dropping liquid device to drip into the reaction vessel with solution, the sealed mouth 15-2 surface of defeated appearance has a layer sealing washer, can combine to play sealed isolated outside air's effect with defeated appearance mouth 11-3, infusion set can be transferred the water sample to the reaction vessel from storage bottle 2 through defeated appearance sealed mouth 15-2 simultaneously, aspiration port 15-4 connect an aspiration pump 31, at first take out the air in the reaction vessel before the chemical operation, make and keep a vacuum state in the reaction vessel, the mouth 15-5 that moves liquid carry a certain amount of solution to spectrophotometer 34 through moving liquid solenoid valve 35.

The transfusion part comprises a water outlet electromagnetic valve 4, a liquid exchanger 5 and a sample transfusion device 11, wherein the water outlet electromagnetic valve 4 is provided with two liquid transfusing amounts which respectively control a sample storage bottle 2 and a water storage bottle 3 to a reaction vessel, the liquid exchanger 5 is made of an anti-oxidation and anti-corrosion material, the liquid exchanger 5 can enable the sample storage bottle 2 and the water storage bottle 3 to transfuse to the reaction vessel through the same set of sample transfusion device 11, the sample transfusion device 11 comprises a sample transfusion tube shell 11-1, an extension tube 11-2, a sample transfusion port 11-3 and a sample transfusion tube 11-4, the tail part of the sample transfusion tube 11-4 is connected with the liquid exchanger 5, the sample transfusion tube shell 11-1 is used for ensuring that the sample transfusion device 11 is vertical to a reaction vessel transfusion sealing port 15-2, the extension tube 11-2 extends out towards the gravity direction under the transfusion working state, so that the sample transfusion port 11-3 is in sealing contact with the transfusion sealing port 15-2, the sample conveying pipe 11-4 is discharged after the sample conveying port 11-3 reaches the designated position, so that the sample conveying pipe is in contact with the bottom of the reaction vessel, and a water sample can slowly flow out along the bottom of the reaction vessel to prevent the water sample from splashing and forming bubbles.

The spectrophotometer 34 is used for reading the iodine concentration in the iodine solution after the chemical reaction is completed, and the iodine concentration is used as one of the input items of the neural network. The solution in the reaction cuvette was quantitatively obtained by a pipette solenoid valve 35, and the obtained solution was pipetted into a quartz glass cuvette in a sample cell in the spectrophotometer 34. The obtained iodine amount concentration information is transmitted to the control panel 1 as an input item of the neural network.

The information processing module comprises a shooting module and a control panel 1, the shooting module comprises a darkroom, a voltage stabilizing light source 14 and a camera 16, the darkroom consists of a shading tail plate 28, a shading device 29 and other wall surfaces in one layer of the cabinet body coated with black light-absorbing pigment, the shading device 29 is divided into an upper plate and a lower plate, and the upper plate moves to contact with the first plate after the chemical operation is finished to form a closed darkroom space; the camera 16, the pressure-stabilizing light source 14 and the quartz glass reaction vessel 15 are positioned at the same height in the center in the shooting state, and the distance between the camera 16, the quartz glass reaction vessel 15 and the pressure-stabilizing light source 14 is adjusted to determine a proper distance; the voltage-stabilizing light source 14 is a long-panel screen type light source, and the brightness of the voltage-stabilizing light source can be adjusted through the control panel 1; the camera 16 shoots the quartz glass reaction vessel 15 in the process of dissolving and precipitating through chemical operation, transmits the obtained picture to the processor of the control panel 1, and judges whether the quartz glass reaction vessel is completely dissolved or not through image processing; after the reaction is completed, the camera 16 takes a picture of the quartz glass reaction vessel 15 in a dark room, and transmits the obtained picture to the processor of the control panel 1, and the obtained picture is subjected to image processing to extract the RGB components and convert the RGB components into HSI components, and then the HSI components are input into the neural network trained by the error back propagation method to obtain the concentration of dissolved oxygen. The real-time monitoring of the water quality of the river channel has the characteristics of serious random interference, strong nonlinearity, large time variation and serious hysteresis, and is difficult to establish an accurate mathematical model through mechanism analysis, so that the neural network adopted by the scheme is very suitable for the control problem of a highly nonlinear and seriously uncertain system, can help to control a detection system to obtain better precision, and ensures the timeliness of the system. The error back propagation method adopts a first-order gradient method (steepest descent method), and a specific calculation formula is as follows:

x_p ⁽¹⁾inputting a pth neuron of the layer;

y_koutputting the k neuron output of the output layer;

t_kand outputting the target output of the kth neuron of the output layer.

Input device

Output of

Error function

A three-layer neural network with four inputs and one output is designed, as shown in fig. 10.

Wherein x₁Is a variable of H, x₂Is the S variable, x₃Is a variable of I, x₄Is the iodine concentration. a is the dissolved oxygen content. The weight omega is calculated by adopting a gradient descent method, namely

For the

In other words.

Note the book

Namely, it is

Then

The popularization is as follows:

the working principle of the device is as follows: in the initial state, each electromagnetic valve is in a closed state, and the liquid dropping device is located at the initial position. According to the sampling demand of the actual river course water sample (water samples at different positions and different depths of the river course), the telescopic distance of the multiple sections of pull rods and the number of rotation turns of the water suction pipe rotating disc are set through the control panel, so that the water suction pipe opening reaches the specified position (the multiple sections of pull rods are assumed to be the maximum telescopic distance). When the multi-section pull rod moves and is in a rod extending state, the sampling motor works and rotates to drive the rolling screw rod to rotate so as to drive the sampling slide block to horizontally move at the front end position of the first section rod at the initial stage on the sampling guide rail, at the moment, the sampling motor stops working, meanwhile, the gas claw fixed on the sliding block enters a working state, the two arms of the gas claw swing downwards to be vacuumized and then the first section rod is fixed, the sampling motor works again to drive the sliding block to move so as to pull the first section rod outwards to a designated position, then the gas claw is released from the working state and drives the sampling motor to move reversely to the front end of the second section rod, the same operation as that of moving the first section rod is carried out until the fourth section rod is completely pulled out of the fifth section rod, at which time the maximum sampling distance is reached, in the process of lengthening each stage of the multi-stage pull rod, the water suction pipe turntable in the fifth stage of the pull rod is driven to rotate, when the preset sampling distance is reached, the rotating disc motor of the suction pipe rotates for a preset number of turns to enable the suction pipe to reach the specified depth. The water suction pump starts to work, the water inlet of the sample storage bottle is opened, a water sample starts to enter the sample storage bottle from the water suction pipe until the water inlet of the sample storage bottle continuously overflows for two minutes, the water inlet of the sample storage bottle is closed, the air suction port of the quartz glass reaction vessel is opened, the air suction pump starts to work, the reaction vessel is vacuumized, then the air suction port is closed, the air suction pump stops working, the sample conveying sealing port is opened, the sample conveying device simultaneously acts, the telescopic pipe extends towards the gravity direction, the sample conveying port is in sealing contact with the infusion sealing port, the sample conveying port conveys out the sample conveying pipe after reaching a specified position, the sample conveying electromagnetic valve 1 at the water outlet of the sample storage bottle is opened after the sample conveying pipe is in contact with the bottom of the reaction vessel, the water sample slowly flows out of the reaction vessel and is closed after 50ml of the water sample is discharged, the sample conveying device is reset, and the sample conveying sealing. Open the drip mouth of reaction ware, because the dropping liquid guide rail is at initial state, the dropping liquid device 1 that is equipped with manganese sulfate solution gets the center and lies in same straight line on the direction of gravity with the drip mouth of reaction ware, and dropping liquid slider 1 drives dropping liquid device 1 and slides along the dropping liquid guide rail, and 1 job control of dropping liquid solenoid valve drips a certain amount of manganese sulfate solution to the reaction ware after reacing the assigned position, later dropping liquid solenoid valve 1 closes, and dropping liquid slider 1 drives dropping liquid device 1 and resets. Step motor work drives the clockwise step motor stop work that rotates 90 degrees of dropping liquid guide rail after, and then dropping liquid slider 2 drives dropping liquid device 2 and slides along the dropping liquid guide rail, and 2 job control of dropping liquid solenoid valve after arriving the assigned position drip a certain amount of sodium hydroxide-potassium iodide solution to the reaction vessel in, later dropping liquid solenoid valve 2 closes, and dropping liquid slider 2 drives dropping liquid device 2 and resets. Rock the rim plate and begin work, drive the reaction ware and rock three minutes in certain space range, make it deposit completely, later rim plate stop work, step motor begins work, drive the dropping liquid guide rail and rotate 90 degrees back step motor stop work clockwise, later dropping liquid slider 3 drives dropping liquid device 3 and slides along the dropping liquid guide rail, reach 3 job control of dropping liquid solenoid valve after the assigned position and drip a certain amount of concentrated sulfuric acid solution to the reaction ware in, later dropping liquid solenoid valve 3 closes, dropping liquid slider 3 drives dropping liquid device 3 and resets. The wheel disc is shaken to start working, the reaction vessel is driven to shake for three minutes in a certain space range, so that the precipitate is completely dissolved, after the wheel disc is shaken to stop working, the wheel disc is static for two minutes, the pressure stabilizing light source works, the camera shoots a solution image of the reaction vessel and transmits the solution image into the control panel, whether the precipitate is completely dissolved or not is analyzed after image processing, and if the precipitate is not complete, the dropping action of the dropping device 3 is continuously repeated until the precipitate is completely dissolved. After the dissolution is completed, the steady-pressure light source and the camera are turned off, the pipetting solenoid valve 35 is turned on, the solution in the reaction cuvette is quantitatively obtained through the pipetting solenoid valve 35, and the obtained solution is pipetted into a quartz glass cuvette in a sample cell in the spectrophotometer 34. The obtained iodine amount concentration information is transmitted to the control panel 1 as one of the input items of the neural network. After the test is finished, the liquid transfer electromagnetic valve is closed, the sample conveying sealing port is opened, the sample conveying device simultaneously acts, the telescopic pipe extends towards the gravity direction, the sample conveying port is in sealing contact with the liquid conveying sealing port, the water outlet electromagnetic valve 2 at the water outlet of the water storage bottle is opened, after 200ml of water sample is discharged, the water outlet electromagnetic valve 2 is closed, the sample conveying device resets, and the sample conveying sealing port is closed. Rock the rim plate and move 1 minute again, make the solution mix completely, close and rock the rim plate, step motor begins work, drives the clockwise 90 degrees back step motor stop work of rotating of dropping liquid guide rail, later dropping liquid slider 4 drives dropping liquid device 4 and slides along the dropping liquid guide rail, reachs behind the assigned position 4 job control of dropping liquid solenoid valve and drips a certain amount of starch solution to the reaction vessel in, later dropping liquid solenoid valve 4 closes, dropping liquid slider 4 drives dropping liquid device 4 and resets. The wheel was shaken for an additional 1 minute to allow complete mixing of the solution and the wheel was closed. The shading device acts, and the upper plate moves to contact with the first plate (the first plate is a cabinet body for fixing the camera) to form a closed darkroom space. The stable pressure light source works, the camera quartz glass reaction vessel shoots, the obtained picture is transmitted to a processor of the control panel, the RGB component of the picture is extracted through image processing, the RGB component is converted into the HSI component, and the HSI component is used as an input quantity and is input into the neural network trained through the error back propagation method together with the iodine concentration measured by the spectrophotometer to obtain the dissolved oxygen concentration. The light source and camera are turned off and the drain valve is opened to drain the waste liquid into the waste liquid tank.

The detection method of the dissolved oxygen of the device is obtained according to the national standard iodometry, firstly, a certain amount of manganese sulfate solution is added into a collected water sample, and then a certain amount of sodium hydroxide-potassium iodide solution is added to generate Mn (OH) immediately₂Precipitation due to Mn (OH)₂Unstable and rapidly react with dissolved oxygen in water to form MnMnO₃Precipitation, which partially reacts to fix dissolved oxygen in water to MnMnO₃Exist in the form of (1). After addition of concentrated sulfuric acid, MnMnO₃Reacts with KI in the solution to release free iodine, and the more iodine separated out represents the higher concentration of dissolved oxygen in water. A very small amount of iodine encountered with the polyol starch also immediately formed a dark blue complex. However, since too high iodine concentration causes no color reaction after addition of the starch solution, the solution is diluted to reduce the iodine concentration.

Although the embodiments have been described and illustrated separately, it will be apparent to those skilled in the art that some common techniques may be substituted and integrated between the embodiments, and reference may be made to one of the embodiments not explicitly described, or to another embodiment described.

The foregoing is illustrative of the preferred embodiments of the present invention only and is not to be construed as limiting the claims. The present invention is not limited to the above embodiments, and the specific structure thereof is allowed to vary. In general, all changes which come within the scope of the invention as defined by the independent claims are intended to be embraced therein.