阅读说明：本技术 一种基于人工神经网络软测量技术的加热炉智能燃烧控制系统 (Heating furnace intelligent combustion control system based on artificial neural network soft measurement technology ) 是由 高建 杨志 刘立国 于 2019-09-16 设计创作，主要内容包括：本发明公开了一种基于人工神经网络软测量技术的加热炉智能燃烧控制系统,包括燃料气输入和燃料气在线热值仪,所述燃料气输入的相应变量输出端为燃料气在线热值仪,所述燃料气在线热值仪的输出端设置有两组,所述燃料气在线热值仪的一组输出端为燃料气热值软测量模块,所述燃料气在线热值仪的另一组输出端为燃料气状态检测模块,所述燃料气状态检测模块的输出端为燃料气热值软测量模块,所述燃料气热值软测量模块的输出端为温度控制系统,所述温度控制系统的输出端为热值前馈模块,解决燃料气热值变化干扰对炉内温度的影响,且该控制方法无需通过人工进行介入,可自主进行精确调节,更加智能化,实用性俱佳。(The invention discloses a heating furnace intelligent combustion control system based on artificial neural network soft measurement technology, which comprises a fuel gas input and a fuel gas online heat value meter, wherein the corresponding variable output ends of the fuel gas input are the fuel gas online heat value meters, two groups of output ends of the fuel gas online heat value meters are arranged, one group of output ends of the fuel gas online heat value meters are fuel gas heat value soft measurement modules, the other group of output ends of the fuel gas online heat value meters are fuel gas state detection modules, the output end of the fuel gas state detection module is a fuel gas heat value soft measurement module, the output end of the fuel gas heat value soft measurement module is a temperature control system, the output end of the temperature control system is a heat value feedforward module, the influence of fuel gas heat value change interference on the temperature in a furnace is solved, and the control method does not need manual intervention, can independently carry out accurate regulation, it is more intelligent, the practicality is all good.)

1. The utility model provides a heating furnace intelligence combustion control system based on soft measurement technique of artificial neural network, includes online calorific value appearance (2) of fuel gas input (1) and fuel gas, its characterized in that: the corresponding variable output end of the fuel gas input (1) is a fuel gas online heat value meter (2), two groups of output ends of the fuel gas online heat value instrument (2) are arranged, one group of output ends of the fuel gas online heat value instrument (2) is a fuel gas heat value soft measuring module (3), the other group of output ends of the fuel gas on-line heat value instrument (2) is a fuel gas state detection module (8), the output end of the fuel gas state detection module (8) is a fuel gas heat value soft measurement module (3), the output end of the fuel gas heat value soft measuring module (3) is a temperature control system (4), the output end of the temperature control system (4) is a heat value feedforward module (5), the output end of the heat value feedforward module (5) is a furnace temperature control module (6), the output end of the furnace temperature control module (6) is a furnace temperature regulation and control module (7).

2. The intelligent combustion control system of the heating furnace based on the artificial neural network soft measurement technology of claim 1, characterized in that: fuel gas input (1) includes fuel gas degree of mixing (11), fuel gas composition (12), fuel gas velocity of flow (13), fuel gas system pressure (14) and fuel gas temperature (15), fuel gas degree of mixing (11), fuel gas composition (12), fuel gas velocity of flow (13), fuel gas system pressure (14) and fuel gas temperature (15) are the online calorific value appearance of fuel gas (2) the measured fuel gas variable value, the online calorific value appearance (2) output of corresponding fuel gas can change when the variable of fuel gas degree of mixing (11), fuel gas composition (12), fuel gas velocity of flow (13), fuel gas system pressure (14) and fuel gas temperature (15) appears changing.

3. The intelligent combustion control system of the heating furnace based on the artificial neural network soft measurement technology of claim 1, characterized in that: the fuel gas heat value soft measurement module (3) comprises an artificial neural network (31), the fuel gas heat value soft measurement module (3) is based on the artificial neural network (31), the fuel gas heat value soft measurement module (3) can construct auxiliary variables related to the fuel gas heat value and a fuel gas heat value output training sample, the fuel gas heat value soft measurement module (3) conducts self-adaptive learning on weight parameters of the artificial neural network (31), the target of the fuel gas heat value soft measurement module (3) enables the artificial neural network (31) to output approximately a heat value instrument, and the fuel gas heat value soft measurement module (3) detects the prediction performance of the artificial neural network (31) model on heat value change through a test sample.

4. The intelligent combustion control system of the heating furnace based on the artificial neural network soft measurement technology of claim 1, characterized in that: the fuel gas state detection module (8) comprises measurement interference (81), data defect (82) and data jump (83), the fuel gas state detection module (8) can process the measurement interference (81), the data defect (82) and the data jump (83) output by the analyzer according to the characteristics of the data output by the fuel gas online heat value meter (2) and remove gross error data, and the fuel gas state detection module (8) is connected with the fuel gas heat value soft measurement module (3) and corrects the fuel gas heat value soft measurement module (3) in real time by using normal data.

5. The intelligent combustion control system of the heating furnace based on the artificial neural network soft measurement technology of claim 1, characterized in that: the flow structure formed by the fuel gas online heat value instrument (2), the fuel gas state detection module (8) and the fuel gas heat value soft measurement module (3) is an intelligent learning flow, and the fuel gas heat value soft measurement module (3) can synthesize various data generated by the fuel gas online heat value instrument (2) and the fuel gas state detection module (8) to simulate and output variables.

6. The intelligent combustion control system of the heating furnace based on the artificial neural network soft measurement technology of claim 1, characterized in that: the temperature control system (4) can process the variables through the heat value feedforward module (5) after the fuel gas heat value soft measurement module (3) obtains the corresponding variables and outputs the variables.

7. The intelligent combustion control system of the heating furnace based on the artificial neural network soft measurement technology of claim 1, characterized in that: the calorific value feedforward module (5) carries out time sequence matching on the corrected calorific value output of the fuel gas, the calorific value feedforward module (5) can enable the calorific value change to be consistent with the time sequence of the temperature change in the furnace after the time matching, and the calorific value feedforward module (5) can set the numerical value of an amplification coefficient according to the influence amplitude of the temperature in the furnace.

8. The intelligent combustion control system of the heating furnace based on the artificial neural network soft measurement technology of claim 1, characterized in that: the heat value feedforward module (5), the furnace temperature control module (6) and the furnace temperature regulation and control module (7) are matched with each other through time difference, and the intervention of the heat value feedforward module (5), the furnace temperature control module (6) and the furnace temperature regulation and control module (7) can seal the heat value difference caused by the time difference.

9. The intelligent combustion control system of the heating furnace based on the artificial neural network soft measurement technology of claim 1, characterized in that: an intelligent system consisting of a fuel gas input (1), a fuel gas online heat value meter (2), a fuel gas heat value soft measurement module (3), a temperature control system (4), a heat value feedforward module (5), a furnace temperature control module (6), a furnace temperature regulation and control module (7) and a fuel gas state detection module (8) can be automatically regulated and controlled without manual intervention.

Technical Field

The invention belongs to the related technical field of petrochemical equipment, and particularly relates to an intelligent combustion control system of a heating furnace based on an artificial neural network soft measurement technology.

Background

If the mixing proportion of the fuel gas used by the heating furnace in the production process changes, the heat value of the mixed fuel gas also changes, and cannot be predicted in advance, at present, operators generally judge that the heat value of the fuel gas has changed through indirect parameters such as air flow, furnace temperature, flue gas temperature and the like, and then adjust the air-fuel ratio according to experience, so that the adjustment timeliness is poor, the oxidation burning loss is increased, the heating quality is reduced, and the atmospheric pollution is easily caused.

The existing heating furnace intelligent combustion control system technology has the following problems: at present, for a heating furnace combustion large-lag control system in China, the main control methods comprise a PID regulator control method, a Schmitt estimation compensation method, a fuzzy logic control method, a PI parameter self-learning and fuzzy temperature control combined double-cross amplitude limiting control method and the like, and the problems of low temperature control precision, large fluctuation, slow system response time and the like generally exist in the methods.

Disclosure of Invention

The invention aims to provide an intelligent heating furnace combustion control system based on an artificial neural network soft measurement technology, and aims to solve the problems that the conventional domestic large-hysteresis control system for heating furnace combustion proposed in the background art mainly comprises a PID (proportion integration differentiation) regulator control method, a Schmitt pre-estimation compensation method, a fuzzy logic control method, a PI parameter self-learning and fuzzy temperature control combined double-cross amplitude limiting control method and the like, and the problems of low temperature control precision, large fluctuation, slow system response time and the like generally exist in the method.

In order to achieve the purpose, the invention provides the following technical scheme:

an intelligent combustion control system of a heating furnace based on an artificial neural network soft measurement technology comprises a fuel gas input and a fuel gas online heat value instrument, the output end of the corresponding variable input of the fuel gas is a fuel gas online heat value instrument, two groups of output ends of the fuel gas online heat value instruments are arranged, one group of output ends of the fuel gas online heat value instrument is a fuel gas heat value soft measuring module, the other group of output ends of the fuel gas online heat value instrument is a fuel gas state detecting module, the output end of the fuel gas state detection module is a fuel gas heat value soft measurement module, the output end of the fuel gas heat value soft measurement module is a temperature control system, the output end of the temperature control system is a heat value feedforward module, the output end of the heat value feedforward module is a furnace temperature control module, and the output end of the furnace temperature control module is furnace temperature regulation and control.

Preferably, the fuel gas input comprises a fuel gas mixing degree, a fuel gas composition, a fuel gas flow rate, a fuel gas system pressure and a fuel gas temperature, the fuel gas mixing degree, the fuel gas composition, the fuel gas flow rate, the fuel gas system pressure and the fuel gas temperature are fuel gas variable values measured by an online fuel gas heat value meter, and the corresponding output quantity of the online fuel gas heat value meter changes when the variables of the fuel gas mixing degree, the fuel gas composition, the fuel gas flow rate, the fuel gas system pressure and the fuel gas temperature change.

Preferably, the fuel gas heat value soft measurement module comprises an artificial neural network, the fuel gas heat value soft measurement module is based on the artificial neural network, the fuel gas heat value soft measurement module can construct auxiliary variables related to the fuel gas heat value and a fuel gas heat value output training sample, the fuel gas heat value soft measurement module performs adaptive learning on weight parameters of the artificial neural network, the artificial neural network output approaches to the output of a heat value instrument by the target of the fuel gas heat value soft measurement module, and the fuel gas heat value soft measurement module tests the prediction performance of the artificial neural network model on heat value change through the test sample.

Preferably, the fuel gas state detection module comprises measurement interference, data defect and data jump, the fuel gas state detection module can process the measurement interference, the data defect and the data jump output by the analyzer according to the characteristics of the data output by the fuel gas online heat value meter and remove gross error data, and the fuel gas state detection module is connected with the fuel gas heat value soft measurement module and corrects the fuel gas heat value soft measurement module in real time by using normal data.

Preferably, the flow structure formed by the fuel gas online heat value meter, the fuel gas state detection module and the fuel gas heat value soft measurement module is an intelligent learning flow, and the fuel gas heat value soft measurement module can synthesize various data generated by the fuel gas online heat value meter and the fuel gas state detection module to simulate output variables.

Preferably, the temperature control system can process the variables through the heat value feedforward module after the fuel gas heat value soft measurement module obtains the corresponding variables and outputs the variables.

Preferably, the calorific value feedforward module performs time sequence matching on the corrected calorific value output of the fuel gas, the calorific value feedforward module can enable the calorific value change to be consistent with the time sequence of the temperature change in the furnace after the time matching, and the calorific value feedforward module can set the numerical value of the amplification factor according to the influence amplitude of the temperature in the furnace.

Preferably, the heat value feedforward module, the furnace temperature control module and the furnace temperature regulation and control are matched with each other in time difference, and the intervention of the heat value feedforward module, the furnace temperature control module and the furnace temperature regulation and control can seal the heat value difference caused by the time difference.

Preferably, an intelligent system consisting of a fuel gas input module, a fuel gas online heat value meter, a fuel gas heat value soft measurement module, a temperature control system, a heat value feedforward module, a furnace temperature control module, a furnace temperature regulation and control module and a fuel gas state detection module can be automatically regulated and controlled without manual intervention.

Compared with the prior art, the invention provides an intelligent combustion control system of a heating furnace based on an artificial neural network soft measurement technology, which has the following beneficial effects:

the intelligent combustion control method based on the artificial neural network soft measurement technology is provided, the fuel gas heat value artificial neural network soft measurement technology is adopted, the problem of time lag of output of an online heat value instrument is solved, the fuel gas heat value feedforward control technology is adopted, the influence of change interference of the fuel gas heat value on the temperature in the furnace is solved, manual intervention is not needed, accurate adjustment can be automatically carried out, and the intelligent combustion control method is more intelligent and good in practicability.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention without limiting the invention in which:

FIG. 1 is a schematic structural diagram of an overall process in an intelligent combustion control system of a heating furnace based on an artificial neural network soft measurement technology, provided by the invention;

FIG. 2 is a schematic view of a flow structure of a fuel gas calorific value soft measurement system in an intelligent combustion control system of a heating furnace based on an artificial neural network soft measurement technology according to the present invention;

FIG. 3 is a schematic view of a flow structure of a temperature control system in an intelligent combustion control system of a heating furnace based on an artificial neural network soft measurement technology according to the present invention;

in the figure: 1. inputting fuel gas; 2. a fuel gas online heat value meter; 3. a fuel gas calorific value soft measurement module; 4. a temperature control system; 5. a calorific value feedforward module; 6. a furnace temperature control module; 7. regulating and controlling the temperature in the furnace; 8. a fuel gas state detection module; 11. the degree of mixing of fuel gas; 12. a fuel gas composition; 13. a fuel gas flow rate; 14. fuel gas system pressure; 15. fuel gas temperature; 31. an artificial neural network; 81. measuring interference; 82. data is defective; 83. and (6) data jumping.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Referring to fig. 1-3, the present invention provides a heating furnace intelligent combustion control system based on artificial neural network soft measurement technology, and the technical scheme is as follows:

the utility model provides a heating furnace intelligence combustion control system based on soft measurement technique of artifical neural network, including the online calorific value appearance 2 of fuel gas input 1 and fuel gas, the corresponding variable output of fuel gas input 1 is the online calorific value appearance 2 of fuel gas, fuel gas input 1 includes fuel gas degree of mixing 11, fuel gas composition 12, fuel gas velocity of flow 13, fuel gas system pressure 14 and fuel gas temperature 15, fuel gas degree of mixing 11, fuel gas composition 12, fuel gas velocity of flow 13, fuel gas system pressure 14 and fuel gas temperature 15 are the online calorific value appearance 2 measured fuel gas variable value of fuel gas, fuel gas degree of mixing 11, fuel gas composition 12, fuel gas velocity of flow 13, the online calorific value appearance 2 output of corresponding fuel gas can change when the variable of fuel gas system pressure 14 and fuel gas temperature 15 appears changing.

An intelligent combustion control system of a heating furnace based on an artificial neural network soft measurement technology comprises two groups of output ends of a fuel gas online heat value instrument 2, wherein one group of output ends of the fuel gas online heat value instrument 2 is a fuel gas heat value soft measurement module 3, the fuel gas heat value soft measurement module 3 comprises an artificial neural network 31, the fuel gas heat value soft measurement module 3 is based on the artificial neural network 31, the fuel gas heat value soft measurement module 3 can construct auxiliary variables related to the fuel gas heat value and a fuel gas heat value output training sample, the fuel gas heat value soft measurement module 3 carries out self-adaptive learning on weight parameters of the artificial neural network 31, the target of the fuel gas heat value soft measurement module 3 enables the artificial neural network 31 to output approximate to the output of a heat value instrument, and the fuel gas heat value soft measurement module 3 tests the prediction performance of the artificial neural network 31 model on, the flow structure formed by the fuel gas online heat value meter 2, the fuel gas state detection module 8 and the fuel gas heat value soft measurement module 3 is an intelligent learning flow, and the fuel gas heat value soft measurement module 3 can synthesize various data generated by the fuel gas online heat value meter 2 and the fuel gas state detection module 8 to simulate and output variables.

The utility model provides a heating furnace intelligence combustion control system based on artifical neural network soft measurement technique, another group's output including the online calorific value appearance 2 of fuel gas is fuel gas state detection module 8, fuel gas state detection module 8's output is fuel gas calorific value soft measurement module 3, fuel gas state detection module 8 is including measuring interference 81, data defect 82 and data jump 83, fuel gas state detection module 8 can be based on the characteristics of the online calorific value appearance 2 output data of fuel gas and come measuring interference 81 to the analyzer output, data defect 82 and data jump 83 handle and get rid of the gross error data, fuel gas state detection module 8 connects fuel gas calorific value soft measurement module 3 and carries out real-time correction with normal data to fuel gas calorific value soft measurement module 3.

The utility model provides a heating furnace intelligence combustion control system based on artifical neural network soft measurement technique, the output that includes fuel gas calorific value soft measurement module 3 is temperature control system 4, the output of temperature control system 4 is calorific value feedforward module 5, fuel gas calorific value soft measurement module 3 derives corresponding variable and exports back temperature control system 4 can handle the variable through calorific value feedforward module 5, calorific value feedforward module 5 carries out time series matching to the fuel gas calorific value output after the correction, calorific value feedforward module 5 can make the calorific value change unanimous with the interior temperature variation chronogenesis of stove after the time matching, calorific value feedforward module 5 can set up the numerical value of amplification factor according to the influence range of stove temperature.

The utility model provides a heating furnace intelligence combustion control system based on artifical neural network soft measurement technique, the output that includes calorific value feedforward module 5 is stove temperature control module 6, the output of stove temperature control module 6 is stove temperature regulation and control 7, calorific value feedforward module 5, stove temperature control module 6 and stove temperature regulation and control 7 are time difference phase-match flow, calorific value feedforward module 5, the intervention of stove temperature control module 6 and stove temperature regulation and control 7 can seal the calorific value difference that the time difference brought, fuel gas input 1, online calorific value appearance 2 of fuel gas, fuel gas calorific value soft measurement module 3, temperature control system 4, calorific value feedforward module 5, stove temperature control module 6, the intelligent system that stove temperature regulation and control 7 and fuel gas state detection module 8 constitute need not artifical intervention can be automatic to regulate and control.

The working principle and the using process of the invention are as follows: after the invention is installed, the mixing degree of fuel gas 11, the components of fuel gas 12, the flow velocity of fuel gas 13, the pressure of fuel gas system 14 and the temperature of fuel gas 15 form the variable of fuel gas input 1, the variable is detected by the fuel gas on-line heat value meter 2, the sensitivity of the variable and the information of the heat value of the fuel gas is analyzed after the detection, then the change of the heat value of the fuel gas is calculated in real time according to the auxiliary variable influencing the heat value of the fuel gas by constructing a bottom fuel gas heat value soft measurement model, meanwhile, an on-line correction system of a fuel gas heat value soft measurement module 3 is constructed, the output of the on-line heat value meter 2 of the fuel gas is used for correcting the output of the soft measurement model, meanwhile, the fuel gas heat value soft measurement module 3 constructs the auxiliary variable relevant to the heat value of the fuel gas, the target makes the artificial neural network 31 output approach the output of the heat value instrument, after training, the prediction performance of the artificial neural network 31 model to the heat value change is checked through a test sample, in the prediction process, variables such as measurement interference 81, data defect 82 and data jump 83 are processed through a fuel gas state detection module 8, gross error data are removed after processing, the soft measurement output of the heat value is corrected in real time by normal data, the corrected variables are online estimated to the heat value of the fuel gas through a data consistency correction technology in a temperature control system 4 and the neural network, further, the feed-forward real-time inference control of the heat value of the furnace is realized, in the control process, a heat value feed-forward module 5 carries out time sequence matching on the corrected heat value output of the fuel gas, the time sequence of the heat value change is consistent with the time sequence of the temperature change in the furnace, and simultaneously, according to the, and setting the numerical value of the amplification factor, and then stably controlling the temperature in the furnace through a furnace temperature control module 6 and a furnace temperature regulation and control module 7.

Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.