阅读说明：本技术 一种多冗余不确定测量参数的控制信号优选方法 (Control signal optimization method for multiple redundant uncertain measurement parameters ) 是由 魏勇 江学文 李楠 叶君辉 周晓亮 赵虹 于 2021-08-09 设计创作，主要内容包括：本发明一种多冗余不确定测量参数的控制信号优选方法,解决现有技术中系统定位检测存在偏移,需要出现较大偏差才进行人工切换,精确的定位系统不能及时投入运行,导致运行效果差,甚至产生运行风险的问题。其步骤包括构建参数向量,根据参数向量构建矩阵B,根据矩阵B构建时间序列,判断当前周期时间序列是否存在异常,对异常参数进行剔除,对矩阵每行进行平方和后选取出最优值,获取对应参数作为系统当前主控信号。本发明通过历史统计分析方法自动识别当前主备控制信号中的异常值并进行剔除；当主用控制信号出现异常或较大偏差时,能够自动选取更优控制信号进行自动切换,提高了全自动运行的控制效果。(The invention discloses a control signal optimization method for multiple redundant uncertain measurement parameters, which solves the problems that in the prior art, the system positioning detection has deviation, the manual switching is carried out only when a large deviation occurs, and an accurate positioning system cannot be put into operation in time, so that the operation effect is poor, and even the operation risk is generated. The method comprises the steps of constructing a parameter vector, constructing a matrix B according to the parameter vector, constructing a time sequence according to the matrix B, judging whether the time sequence of the current period is abnormal or not, eliminating abnormal parameters, performing square sum post-selection on each row of the matrix to obtain an optimal value, and acquiring corresponding parameters as current main control signals of the system. The invention automatically identifies abnormal values in the current main/standby control signals and eliminates the abnormal values by a historical statistical analysis method; when the main control signal is abnormal or has larger deviation, a more optimal control signal can be automatically selected for automatic switching, and the control effect of full-automatic operation is improved.)

1. A control signal optimization method for multiple redundant uncertain measurement parameters is characterized by comprising the following steps: the method comprises the following steps:

s1, detecting the redundant equipment data of the system in real time, detecting at the same time to obtain n multiple redundant uncertain parameters, and constructing multiple redundant uncertain measurement parameter vectors { a_i，i＝1，2，…n}；

S2, constructing an n multiplied by n matrix according to the uncertain measurement parameter vector

Wherein b is_ij＝a_i–a_j，b_ii＝0，i＝1，2，…n，j＝1，2，…n；

S3, according to the periodic characteristics of system operation and parameter detection, b in the past time is constructed_ijTime series b of_ij(k) I ≠ j, k ≠ 1, 2, … m, where k is the number of cycles and m is satisfied b_ij(k) Total number of sequences normally distributed;

s4, calculating b_ij(k) Mean value ofSum root mean square difference delta_ijWhere i ≠ j, k ≠ 1, 2 … m;

s5, determining the current period { b } in the loop i, i ═ 1, 2, … n_ij(0) Whether j is 1, 2, … n, i ≠ j } all satisfiesIf yes, the corresponding parameter a_iRecording the abnormal data, and entering step S6, if not, continuing to circulate, and entering step S7 after the circulation is finished;

s6, never determining the measurement parameter vector { a_i1, 2, … n } eliminating abnormal parameter a_iObtaining a new uncertain measurement parameter vector { a_i，i＝1，2，…n₁N is constructed by using a new uncertain measurement parameter vector₁×n₁Matrix B₁；

And S7, calculating the sum of squares of each row of data in the finally obtained matrix, selecting a minimum value from the sum of squares, finding a parameter corresponding to the minimum value of the sum of squares, and selecting the parameter as a current main control signal of the system.

2. The method for optimizing control signals for multiple redundant uncertain measured parameters according to claim 1, wherein the step S3 is to set the time sequence period T according to the periodic characteristics of system operation and parameter detection to construct b_ij(k) I ≠ j, k ≠ 1, 2, … m is the average of multiple measurements in the kth cycle in the past, b_ij(0) Is the average of multiple measurements over the current period.

3. The method for controlling signal optimization of multiple redundant uncertain measurement parameters of claim 1, wherein the step S7 comprises the following steps:

s701, when the finally obtained matrix is a matrix B, calculating the square sum of each row of data of the matrix B,

s702, obtaining D_min(0)＝min{D_i(0) I ═ 1, 2, … n }, find D_min(0) Corresponding parameter a_iAnd selecting the parameter as a current main control signal of the system.

4. The method for controlling signal optimization of multiple redundant uncertain measurement parameters of claim 1, wherein the step S7 comprises the following steps:

s711, when the finally obtained matrix is a matrix B₁Calculating the matrix B₁The sum of the squares of each row of data,

s712, obtaining D_min(0)＝min{D_i(0)，i＝1，2，…n₁Find D_min(0) Corresponding parameter a_iAnd selecting the parameter as a current main control signal of the system.

5. The method for controlling signals of multiple redundant uncertain measured parameters according to claims 3 or 4, characterized in that step S7 is preceded by the following steps:

s71, judging n or n according to the finally obtained matrix₁If the value is greater than 2, the process proceeds to step S7, otherwise, the process proceeds to the next step;

s72, judging n or n₁Whether the value is equal to 2 or not, if so, the parameter a corresponding to any data in the matrix_iAs the current main control signal of the system, if not, entering the next step;

s73, judging n or n₁Whether the matrix is equal to 1 or not, if so, the parameter a corresponding to the data in the matrix_iThe system can not be used as the current main control signal of the system, and the system stops running.

6. The method of claim 1, further comprising a preprocessing step before step S1, wherein the preprocessing step comprises:

and detecting the redundant equipment data of the system in real time, and when the system is detected to have a fault, rejecting the parameters of the fault equipment by detecting a fault signal of the system equipment, and entering the step S1.

Technical Field

The invention relates to the technical field of electric digital data processing, in particular to a control signal optimization method for multiple redundant uncertain measurement parameters.

Background

The bucket wheel machine is the key operation equipment of bulk cargo field bulk cargo pier, and domestic full automatization operation prevalence is lower, is in full-automatic examination point or the preliminary stage of popularization at present. The full-automatic starting point and key technology for realizing the bucket wheel machine lie in the accurate positioning of an operation point, in order to improve the positioning reliability, most manufacturers adopt a dual-redundancy or multi-redundancy technology, and adopt 3 sets or more sets of detection equipment respectively on cart positioning, cantilever rotation angle and cantilever pitching angle detection. In an open field, a Beidou precise positioning technology and a high-precision encoder technology redundancy mode is usually adopted, and a plurality of sets of high-precision encoder technology redundancy modes are adopted in a closed field. For the positioning detection of redundant configuration, a full-automatic control system generally adopts a fixed priority assigned system as a main system, one or more other systems as standby systems, the standby systems simultaneously check the availability of the main system in real time, and when the main system and the standby systems have larger deviation, the system automatically stops working and is manually analyzed and switched to a new available positioning system or an original positioning system is reserved.

The existing method has some defects, because the redundant positioning detection is adopted, the main and standby positioning systems are fixedly and preferentially assigned, when the main positioning system deviates due to accumulated errors, the main positioning system cannot be actively switched to the standby system, and the manual switching is required to be carried out when large deviation occurs between the main and standby positioning systems, so that the full-automatic control system can operate for a long time when the positioning system with certain deviation exists, the real more accurate positioning system cannot be put into operation in time, and as a result, the full-automatic control operation effect of the bucket wheel machine is poor, and even the operation safety risk is generated. For example, a bucket wheel machine rotation angle encoder detects that the bucket wheel machine runs on the same side of a track for a long time, accumulated errors are gradually generated, the maximum error is 2-3 degrees, and for a bucket wheel machine with the cantilever radius R being 50m, the position deviation of 2.62m at the maximum is generated in coal yard operation, the coal piling and taking effect is greatly reduced, and even the bucket wheel machine possibly collides with a cart track foundation or an outer side coal retaining wall, so that production safety accidents are caused.

Disclosure of Invention

The invention mainly solves the problems that in the prior art, the system positioning detection has deviation, the manual switching is carried out only when a large deviation occurs, and an accurate positioning system cannot be put into operation in time, so that the operation effect is poor, and even the operation risk is generated, and provides a control signal optimization method of multiple redundant uncertain measurement parameters.

The technical problem of the invention is mainly solved by the following technical scheme: a control signal optimization method for multiple redundant uncertain measurement parameters is characterized by comprising the following steps: the method comprises the following steps:

s1, detecting the redundant equipment data of the system in real time, detecting at the same time to obtain n multiple redundant uncertain parameters, and constructing multiple redundant uncertain measurement parameter vectors { a_i，i＝1，2，…n}；

S2, constructing an n multiplied by n matrix according to the uncertain measurement parameter vector

Wherein b is_ij＝b_i–b_j，b_ii＝0，i＝1，2，…n，j＝1， 2，…n；

S3, according to the periodic characteristics of system operation and parameter detection, b in the past time is constructed_ijTime series b of_ij(k) I ≠ j, k ≠ 1, 2, … m, where k is the number of cycles and m is satisfied b_ij(k) Total number of sequences normally distributed;

s4, calculating b_ij(k) Mean value ofSum root mean square difference delta_ijWhere i ≠ j, k ≠ 1, 2 … m; here for b_ij(k) And k is 1, 2, … m historical data, and the average value and the root mean square difference are calculated.

S5, determining the current period { b } in the loop i, i ═ 1, 2, … n_ij(0) Whether j is 1, 2, … n, i ≠ j } all satisfiesIf yes, the corresponding parameter a_iRecording the abnormal data, entering step S6, if not, continuing to circulate, and entering step after the circulation is finishedA step S7; in the step, the average value and the root mean square deviation time sequence b are calculated according to historical data_ij(k) The judgment of whether there is an abnormality can reduce the probability that the parameter in normal operation is misjudged to 0.27%. In the scheme, if the loop i is currently set as b_ij(0) If j is 1, 2, … n is abnormal data, then b is currently_ij(0) The parameters corresponding to j 1, 2, … n are recorded as abnormal parameters, all the abnormal parameters are found through the loop i, and the process goes to step S6. If it is present b_ij(0) If there is no abnormal data, the process proceeds to step S7 after the cycle is completed.

S6, never determining the measurement parameter vector { a_i1, 2, … n } eliminating abnormal parameter a_iObtaining a new uncertain measurement parameter vector { a_i，i＝1，2，…n₁N is constructed by using a new uncertain measurement parameter vector₁×n₁Matrix B₁(ii) a This step is used to reject the abnormal parameters.

And S7, calculating the sum of squares of each row of data in the finally obtained matrix, selecting a minimum value from the sum of squares, finding a parameter corresponding to the minimum value of the sum of squares, and selecting the parameter as a current main control signal of the system. After the abnormity judgment, the matrix obtained without abnormal parameters is the matrix B, and the matrix B is obtained after the abnormal data is deleted₁。

The invention automatically identifies abnormal values in the current main and standby control signals and eliminates the abnormal values through a historical statistical analysis method, and can automatically select a more optimal control signal to automatically switch when the main control signal is abnormal or has larger deviation, thereby improving the control effect of full-automatic operation.

Preferably, in step S3, the time-series period T is set according to the periodic characteristics of the system operation and the parameter detection, and b is constructed_ij(k) I ≠ j, k ≠ 1, 2, … m is the average of multiple measurements in the kth cycle in the past, b_ij(0) Is the average of multiple measurements over the current period. When the system is in a non-operation state (which means that the detection system redundant equipment is possibly in an operation state, but the detected equipment is not in operation), the measurement value of the detection system redundant equipment is invalidThe value is that the time period is an invalid period, and other time periods are valid periods; and selecting the latest m periods as a sequence from effective periods which extend from the current period to the time when the redundant equipment of the detection system is manually re-calibrated. Wherein m is not less than 100, or m<The actual number of active cycles is taken at 100.

As a preferable scheme, the specific process of step S7 includes:

s701, when the finally obtained matrix is a matrix B, calculating the square sum of each row of data of the matrix B,

s702, obtaining D_min(0)＝min{D_i(0) I ═ 1, 2, … n }, find D_min(0) Corresponding parameter a_iAnd selecting the parameter as a current main control signal of the system.

As a preferable scheme, the specific process of step S7 includes:

s711, when the finally obtained matrix is a matrix B₁Calculating the matrix B₁The sum of the squares of each row of data,

s712, obtaining D_min(0)＝min{D_i(0)，i＝1，2，…n₁Find D_min(0) Corresponding parameter a_iAnd selecting the parameter as a current main control signal of the system.

As a preferable scheme, step S7 is preceded by the following steps:

s71, judging n or n according to the finally obtained matrix₁If the value is greater than 2, the process proceeds to step S7, otherwise, the process proceeds to the next step; after the anomaly judgment, if the matrix obtained by the anomaly-free data is a matrix B, judging n; if abnormal data exist, deleting the abnormal data to obtain a matrix B₁Then, to n₁And (6) judging.

S72, judging n orn₁Whether the value is equal to 2 or not, if so, the parameter a corresponding to any data in the matrix_iAs the current main control signal of the system, if not, entering the next step; when n is₁And when the number of the parameters is equal to 2, only two parameters are needed, any one parameter is selected as the main control signal of the new time period, and the other one is used as the standby control signal.

S73, judging n or n₁Whether the matrix is equal to 1 or not, if so, the parameter a corresponding to the data in the matrix_iThe system can not be used as the current main control signal of the system, and the system stops running. When n is₁When the number of the parameters is equal to 1, only one parameter works normally, but specific parameters cannot be determined, the system needs to stop running, manually calibrate the parameters and confirm normal working parameters.

As a preferable scheme, a preprocessing step is further included before step S1, and the process is as follows:

and detecting the redundant equipment data of the system in real time, and when the system is detected to have a fault, rejecting the parameters of the fault equipment by detecting a fault signal of the system equipment, and entering the step S1. When the equipment fails, the parameters of the equipment with the failure are removed, and then the subsequent steps are carried out.

Therefore, the invention has the advantages that:

1. automatically identifying abnormal values in the current main/standby control signals by a historical statistical analysis method and removing the abnormal values;

2. when the main control signal is abnormal or has larger deviation, a more optimal control signal can be automatically selected for automatic switching, and the control effect of full-automatic operation is improved.

Detailed Description

The technical scheme of the invention is further specifically described by the following embodiments.

Example (b):

the embodiment provides a control signal optimization method for multiple redundant uncertain measurement parameters, which comprises the following steps:

pretreatment, the process is as follows:

detecting system redundant equipment data in real time, and when detecting that a system has a fault, rejecting parameters of the fault equipment by detecting a fault signal of the system equipment; the process advances to step S1.

S1, detecting the redundant equipment data of the system in real time, detecting at the same time to obtain n multiple redundant uncertain parameters, and constructing multiple redundant uncertain measurement parameter vectors { a_i，i＝1，2，…n}；

S2, constructing an n multiplied by n matrix according to the uncertain measurement parameter vector

Wherein b is_ij＝b_i-b_j，b_ii0, i-1, 2, … n, j-1, 2, … n; i.e. b_ijIs a_iAnd all a_jThe difference between the values.

S3, setting a time sequence period T according to the periodic characteristics of system operation and parameter detection, and constructing b in the past time_ijTime series b of_ij(k) I ≠ j, k ≠ 1, 2, … m, where k is the number of cycles and m is satisfied b_ij(k) Total number of sequences normally distributed; b_ij(k) Average of a number of measurements over the k-th period in the past, b_ij(0) Is the average of multiple measurements over the current period.

S4, calculating b_ij(k) Mean value ofSum root mean square difference delta_ijWhere i ≠ j, k ≠ 1, 2 … m;

s5, determining the current period { b } in the loop i, i ═ 1, 2, … n_ij(0) Whether j is 1, 2, … n, i ≠ j } all satisfiesIf yes, the corresponding parameter a_iRecording the abnormal data, and entering step S6, if not, continuing to circulate, and entering step S7 after the circulation is finished;

s6, never determining the measurement parameter vector { a_i1, 2, … n } eliminating abnormal parameter a_iObtaining a new uncertain measurement parameter vector { a_i，i＝1，2，…n₁N is constructed by using a new uncertain measurement parameter vector₁×n₁Matrix B₁。

In the previous step S5, no abnormal parameter is detected, the operation is performed on the matrix B after the step S7 is directly performed, and if the abnormal parameter is detected, the operation is performed on the matrix B constructed after the abnormal data are removed after the step S7 is performed₁The method further comprises the following steps before step S7:

s71, according to the finally obtained matrix, taking the matrix as a matrix B or a matrix B₁Judging n or n₁If the value is greater than 2, the process proceeds to step S7, otherwise, the process proceeds to the next step;

s72, judging n or n₁Whether the value is equal to 2 or not, if so, the parameter a corresponding to any data in the matrix_iAs the current main control signal of the system, if not, entering the next step;

s73, judging n or n₁Whether the matrix is equal to 1 or not, if so, the parameter a corresponding to the data in the matrix_iThe system can not be used as the current main control signal of the system, and the system stops running. When n is₁When the number of the parameters is equal to 1, only one parameter works normally, but specific parameters cannot be determined, the system needs to stop running, manually calibrate the parameters and confirm normal working parameters.

And S7, calculating the sum of squares of each row of data in the finally obtained matrix, selecting a minimum value from the sum of squares, finding a parameter corresponding to the minimum value of the sum of squares, and selecting the parameter as a current main control signal of the system.

When no abnormal parameter is detected in step S5, the process proceeds to step S7 to operate the matrix B, and the specific process includes:

s701, when the finally obtained matrix is a matrix B, calculating the square sum of each row of data of the matrix B,

s702, obtaining D_min(0)＝min{D_i(0) I ═ 1, 2, … n }, find D_min(0) Corresponding parameter a_iAnd selecting the parameter as a current main control signal of the system.

When the abnormal data is detected in step S5, the abnormal data is eliminated to form a matrix B₁The process proceeds to step S7 for matrix B₁The operation is carried out, and the specific process comprises the following steps:

s711, when the finally obtained matrix is a matrix B₁Calculating the matrix B₁The sum of the squares of each row of data,

s712, obtaining D_min(0)＝min{D_i(0)，i＝1，2，…n₁Find D_min(0) Corresponding parameter a_iAnd selecting the parameter as a current main control signal of the system.

This embodiment illustrates the method as a specific example. When the bucket wheel machine is transformed in a full-automatic control mode, mutually independent redundant positioning devices such as a Beidou/GPS (global positioning system), an encoder and the like are arranged, and the cart position, the cantilever pitching angle and the cantilever rotation angle of the real-time detection equipment are positioned.

Taking the cart position as an example, two sets of cart encoders which adopt different implementation modes and are mutually independent and a set of Beidou/GPS precise positioning system are arranged, and the Beidou/GPS precise positioning system data is converted into the cart position, the cantilever pitching angle and the cantilever rotating angle through the geometric relation of the bucket wheel machine cart and the cantilever. Because when the bucket wheel machine cart moves back and forth for a long time in a section of track interval, the cart encoder is easy to generate accumulated errors, and the accumulated errors become larger along with the prolonging of time, but the Beidou/GPS precise positioning system is also influenced by celestial bodies and weather environmental factors and has random errors. The method is used to explain how to select the optimal positioning signal as the main control parameter.

Let the measurement parameters of the cart position be a₁，a₂，a₃And the measured values respectively correspond to a cart encoder 1, a cart encoder 2 and a Beidou/GPS precise positioning system.

Real-time detection system of the sameObtaining the 3 multiple redundant uncertain parameters at any moment, and constructing a multiple redundant uncertain measurement parameter vector { a₁，a₂，a₃}。

From an uncertain measurement parameter vector a₁，a₂，a₃Construction of a 3X 3 matrix

Wherein b is_ij＝a_i–a_j，b_ii＝0；

According to the periodic characteristics of system operation and parameter detection, constructing a time sequence period T and constructing b in the past time_ijTime series b of_ij(k) I ≠ j, k ≠ 1, 2, … 100; wherein b is_ij(0) Is t [ -5m, 0 ]]Average value over time period, b_ij(k) Is t [ -5m-kh, -kh]Average over time period, where m is minutes, h is hours, and the period is 1 h. The period of time that the bucket wheel machine stops operating is not in the time series.

Statistical time series b_ij(k) I ≠ j, k ≠ 1, 2, … 100, mean of 100 historical dataSum root mean square difference delta_ij；

Judgment b_ij(0) If i ≠ j is an abnormal value, the judgment procedure is to determine whether it is satisfiedIf yes, the current period b_ij(0) And recording the data as abnormal data, otherwise, calculating the sum of squares by using a matrix B.

For i, i is 1, 2, 3, if all b_ij(0) If j is 1, 2, … n, i ≠ j, it is determined as an abnormal value, the corresponding parameter a_iFor exception, the parameter a is set_iRemoving the undetermined measurement parameter vectors;

using culling exception parameter a_iThe latter parameter vector { a }_i，i＝1，…n₁}, constructing a new n₁×n₁Matrix B₁(ii) a In a matrix B₁A sum of squares calculation is performed.

For n or n₁If the number of the parameters is greater than 2, the row judgment is performed, in this example, the number of the parameters is 3, and if there is no abnormal parameter, the matrix B, where n is 3, is greater than 2, and the sum of squares of the rows of the matrix B is calculated. And if there is an abnormal parameter, the matrix B₁N of (A) to (B)₁Is less than or equal to 2.

The sum of the squares of the rows of matrix B is calculated as follows,

D₁(0)＝b₁₂(0)²+b₁₃(0)²；

D₂(0)＝b₂₁(0)²+b₂₃(0)²；

D₃(0)＝b₃₁(0)²+b₃₂(0)²；

obtaining D_min(0)＝min{D₁(0)，D₂(0)，D₃(0) Get D_min(0) Corresponding parameter a_iThe system selects this parameter as the current master control signal for the system.

If abnormal parameters exist, a new matrix B is constructed after the abnormal parameters are eliminated₁，

If n is₁Equal to 2, then B in the matrix₁Parameter a corresponding to any one of the data_iThe other one is used as a standby control signal as a main control signal of a new time period of the system.

If n is₁Equal to 1, only one parameter is working normally, but no specific parameter, matrix B, can be determined₁Parameter a corresponding to the middle data_iThe system can not be used as the current main control signal of the system, the system stops running, parameters are calibrated manually, and normal working parameters are confirmed.

The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.