Sensor configuration method of mine digital twin system based on multi-task optimization algorithm

Abstract

The application relates to a mine digital twin system sensor configuration method based on a multi-task optimization algorithm, and comprises the following steps: determining a sensor configuration task, the sensor configuration task comprises multiple subtasks, each subtask corresponds to a sensor type; determining an optimization target of each subtask; a variable in the optimization target is a sensor configuration scheme corresponding to the sensor type; a multi-task optimization algorithm is used to solve the optimization target, and an optimal sensor configuration scheme corresponding to each sensor type is determined. According to the application, the mine digital twin system sensor configuration with various sensor types is decomposed according to the sensor type and solved by means of the multi-task optimization algorithm, the configuration effect of the configuration scheme on each type of sensor is improved, and the problems of uneven configuration of different types of sensors and possible redundancy between different sensors can be effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mine sensor configuration, in particular to a mine digital twin system sensor configuration method based on a multi-task optimization algorithm. BACKGROUND

[0002] At present, the digital twin system is combined with the mine digital construction more and more closely, and it becomes a trend to use the digital twin technology to solve the management and early warning analysis problems of the mine. The digital twin system is a real-time dynamic hyper-reality mapping of a physical entity system, and the real-time collection, transmission and updating of sensor data are crucial for the digital twin. The mine digital twin system involves many types of sensors, and needs to monitor various useful and harmful gases, temperature and humidity parameters, and working conditions such as coal mining, tunneling, transportation and ventilation. The good or bad of the sensor configuration scheme is directly related to the diagnosability of the mine digital twin system to various disaster events in the mine, so the sensor configuration is the basis of the entire digital twin system. The existing sensor configuration method of the digital twin system has the problems of unbalanced sensor configuration and possible redundancy between different sensors, which not only causes economic waste, but also leads to insufficient accuracy of sensor data, and cannot accurately monitor each location in the mine. These problems are the intrinsic reasons for the difficulty in quickly and reliably detecting disasters in the mine. SUMMARY

[0003] In order to overcome at least one deficiency in the prior art, the embodiments of the present application provide a mine digital twin system sensor configuration method based on a multi-task optimization algorithm.

[0004] In a first aspect, a mine digital twin system sensor configuration method based on a multi-task optimization algorithm is provided, comprising:

[0005] Step S1, determining a sensor configuration task, the sensor configuration task including a plurality of subtasks, each subtask corresponding to a sensor type;

[0006] Step S2, determining an optimization objective of each subtask; the variable in the optimization objective is a sensor configuration scheme corresponding to the sensor type;

[0007] Step S3, using a multi-task optimization algorithm to solve the optimization objective, and determining an optimal sensor configuration scheme corresponding to each sensor type.

[0008] In one embodiment, using a multi-task optimization algorithm to solve the optimization objective and determining an optimal sensor configuration scheme corresponding to each sensor type includes:

[0009] In step S31, according to the sensor configuration rule, N sensor overall configuration schemes are generated as an initial configuration scheme set, i.e., an initial population, each of which includes sensor configuration schemes corresponding to all sensor types;

[0010] In step S32, if the current population is the initial population, for each sensor overall configuration scheme in the initial population, an optimization target value corresponding to each sensor type is calculated based on the sensor configuration scheme corresponding to each sensor type, and the best-performing sensor type in each sensor overall configuration scheme is recorded according to the optimization target value; if the current population is not the initial population, an optimization target value corresponding to the best-performing sensor type in each sensor overall configuration scheme recorded in the current population is calculated;

[0011] In step S33, according to the optimization target value corresponding to the best-performing sensor type in each sensor overall configuration scheme in the current population, a scheme fitness of each sensor overall configuration scheme is calculated.

[0012] In step S34, the scheme fitnesses are sorted in descending order, and the N sensor overall configuration schemes corresponding to the top N scheme fitnesses are selected as the next population; if the current optimization target calculation frequency FE reaches a set number, the process ends, and the optimal sensor configuration scheme corresponding to each sensor type is determined according to the selected N sensor overall configuration schemes; if the current optimization target calculation frequency FE does not reach the set number, offspring are generated using a crossover operation based on the next population, the best-performing sensor type in the offspring is determined, and each individual in the offspring and the next population is subjected to a mutation operation to obtain mutated individuals as the current population for the next calculation, and the process returns to step S32.

[0013] In one embodiment, the offspring are generated using a crossover operation based on the next population, including:

[0014] Two sensor overall configuration schemes are randomly selected from the N sensor overall configuration schemes in the next population, denoted as a and b;

[0015] If the best-performing sensor types in the sensor overall configuration scheme a and the sensor overall configuration scheme b are the same, the sensor overall configuration scheme a and the sensor overall configuration scheme b are crossed using a simulated binary crossover method to generate two offspring, i.e., a sensor overall configuration scheme c and a sensor overall configuration scheme d, otherwise, the sensor overall configuration scheme a and the sensor overall configuration scheme b are crossed according to a crossover probability to generate two offspring, i.e., a sensor overall configuration scheme c and a sensor overall configuration scheme d, wherein The above operations are performed times.

[0016] ​In one embodiment, determining the best-performing sensor type in the offspring includes:

[0017] 2 offspring are sensor overall configuration scheme c and sensor overall configuration scheme d; for sensor overall configuration scheme c and sensor overall configuration scheme d, two random numbers rand1, rand2 in the range of 0-1 are generated respectively; if rand1<0.5, the best-performing sensor type in sensor overall configuration scheme c is determined as the best-performing sensor type recorded in sensor overall configuration scheme a, and if rand1≥0.5, the best-performing sensor type in sensor overall configuration scheme c is determined as the best-performing sensor type recorded in sensor overall configuration scheme b; if rand2<0.5, the best-performing sensor type in sensor overall configuration scheme d is determined as the best-performing sensor type recorded in sensor overall configuration scheme a, and if rand2≥0.5, the best-performing sensor type in sensor overall configuration scheme d is determined as the best-performing sensor type recorded in sensor overall configuration scheme b.

[0018] In one embodiment, determining the optimal sensor configuration scheme corresponding to each sensor type according to the selected top N sensor overall configuration schemes includes:

[0019] For each sensor type, the sensor configuration scheme with the optimal optimization target value corresponding to the sensor type in the N sensor overall configuration schemes is selected as the optimal sensor configuration scheme corresponding to the sensor type.

[0020] In a second aspect, a sensor configuration device for a mine digital twin system based on a multi-task optimization algorithm is provided, including:

[0021] A sensor configuration task determination module is configured to determine a sensor configuration task, and the sensor configuration task includes a plurality of subtasks, each subtask corresponding to a sensor type;

[0022] An optimization target determination module is configured to determine an optimization target for each subtask; the variable in the optimization target is a sensor configuration scheme corresponding to the sensor type;

[0023] An optimization target solving module is configured to solve the optimization target by using a multi-task optimization algorithm to determine the optimal sensor configuration scheme corresponding to each sensor type.

[0024] In one embodiment, the optimization target solving module is further configured to perform the following functions:

[0025] In step S31, N sensor overall configuration schemes are generated as an initial configuration scheme set, i.e., an initial population, according to a sensor configuration rule, each sensor overall configuration scheme including a sensor configuration scheme corresponding to all sensor types.

[0026] In step S32, if the current population is the initial population, the optimization target value corresponding to each sensor type is calculated based on the sensor configuration scheme corresponding to each sensor type for each sensor overall configuration scheme in the initial population, and the best-performing sensor type in each sensor overall configuration scheme is recorded according to the optimization target value; if the current population is not the initial population, the optimization target value corresponding to the best-performing sensor type in each sensor overall configuration scheme recorded in the current population is calculated.

[0027] In step S33, the scheme fitness of each sensor overall configuration scheme is calculated according to the optimization target value corresponding to the best-performing sensor type in each sensor overall configuration scheme in the current population.

[0028] In step S34, the scheme fitnesses are sorted in descending order, and the sensor overall configuration schemes corresponding to the top N scheme fitnesses are selected as the next population; if the current optimization target calculation frequency FE reaches the set number, the process is ended, and the optimal sensor configuration scheme corresponding to each sensor type is determined according to the selected top N sensor overall configuration schemes; if the current optimization target calculation frequency FE does not reach the set number, the crossover operation is used to generate offspring based on the next population, the best-performing sensor type in the offspring is determined, and the mutation operation is performed on the offspring and each individual in the next population to obtain the mutated individuals as the current population for the next calculation, and the process returns to step S32.

[0029] In one embodiment, the optimization target solving module is further configured to:

[0030] Two sensor overall configuration schemes are randomly selected from the N sensor overall configuration schemes in the next population, denoted as a and b.

[0031] If the best-performing sensor types in the sensor overall configuration scheme a and the sensor overall configuration scheme b are the same, the sensor overall configuration scheme a and the sensor overall configuration scheme b are crossed using the simulated binary crossover method to generate two offspring, i.e., the sensor overall configuration scheme c and the sensor overall configuration scheme d; otherwise, the crossover is performed according to a crossover probability to generate two offspring, i.e., the sensor overall configuration scheme c and the sensor overall configuration scheme d. The above operations are performed for times.

[0032] In one embodiment, the optimization target solving module is further configured to:

[0033] ​​The two offspring are a sensor overall configuration scheme c and a sensor overall configuration scheme d; for the sensor overall configuration scheme c and the sensor overall configuration scheme d, two random numbers rand1, rand2 in the range of 0-1 are generated respectively; if rand1<0.5, the sensor type with the best performance in the sensor overall configuration scheme c is determined as the sensor type with the best performance recorded in the sensor overall configuration scheme a, if rand1 is greater than or equal to 0.5, the sensor type with the best performance in the sensor overall configuration scheme c is determined as the sensor type with the best performance recorded in the sensor overall configuration scheme b; if rand2<0.5, the sensor type with the best performance in the sensor overall configuration scheme d is determined as the sensor type with the best performance recorded in the sensor overall configuration scheme a, if rand2 is greater than or equal to 0.5, the sensor type with the best performance in the sensor overall configuration scheme d is determined as the sensor type with the best performance recorded in the sensor overall configuration scheme b.

[0034] In one embodiment, the optimization target solving module is further configured to:

[0035] For each sensor type, a sensor configuration scheme with the optimal optimization target value corresponding to the sensor type in the N sensor overall configuration schemes is selected as the optimal sensor configuration scheme corresponding to the sensor type.

[0036] Compared with the prior art, the present application has the following beneficial effects: the present application decomposes the sensor configuration of the mine digital twin system with a variety of sensor types according to the sensor type and solves it by means of a multi-task optimization algorithm, improves the configuration effect of the configuration scheme on each type of sensor, and can effectively solve the problems of uneven configuration of different types of sensors and possible redundancy between different sensors. BRIEF DESCRIPTION OF DRAWINGS

[0037] The present application can be better understood by referring to the description given below in conjunction with the accompanying drawings, which are included in the specification and form a part of the specification. In the drawings:

[0038] Figure 1 A flow chart of a mine digital twin system sensor configuration method based on a multi-task optimization algorithm according to an embodiment of the present application is shown;

[0039] Figure 2 A structural block diagram of a mine digital twin system sensor configuration device based on a multi-task optimization algorithm according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0040] The exemplary embodiments of this application will be described hereinafter with reference to the accompanying drawings. In the description, all the features of the practical embodiments are not described for the sake of clarity and conciseness. It should be appreciated that many embodiment-specific decisions can be made in the process of developing any such practical embodiments in order to achieve the specific goals of the developers, and these decisions can vary from embodiment to embodiment.

[0041] It should also be noted here that, in order not to obscure the application due to unnecessary details, only the device structures closely related to the scheme according to the application are shown in the drawings, and other details not closely related to the application are omitted.

[0042] It should be understood that the application is not limited to the described implementation forms due to the following description with reference to the drawings. In this text, the embodiments can be combined with each other, features between different embodiments can be replaced or borrowed, and one or more features can be omitted in one embodiment, if possible.

[0043] The application provides a sensor configuration method of a mine digital twin system based on a multi-task optimization algorithm, and the application scenario is a large coal mine. A mine digital twin system needs to be constructed in the coal mine, and therefore a plurality of sensors need to be deployed to monitor various useful and harmful gases, temperature and humidity and other parameters, and working states such as coal mining, tunneling, transportation and ventilation in the mine. The above monitoring needs to use a plurality of different types of sensors.

[0044] The application embodiment provides a sensor configuration method of a mine digital twin system based on a multi-task optimization algorithm, Figure 1 A flowchart of the sensor configuration method of the mine digital twin system based on the multi-task optimization algorithm according to the application embodiment is shown, and the method comprises the following steps:

[0045] Step S1, determining a sensor configuration task, the sensor configuration task comprising a plurality of subtasks, each subtask corresponding to a sensor type;

[0046] In this step, the sensor type can be set to 10 to meet the measurement of different parameters and the monitoring of different working states. Each subtask is used to determine the configuration quantity and configuration position of the sensor in the sensor type corresponding to the subtask.

[0047] Step S2, determining an optimization objective of each subtask; the variable in the optimization objective is a sensor configuration scheme corresponding to the sensor type; since the evaluation indexes of different types of sensors are different, different optimization objectives need to be set according to the sensor type, and the optimization objectives can include quantity, cost, expected information gain, etc. Here, constraint conditions are set for each optimization objective, and the constraint conditions can include the maximum tolerable cost, the maximum distance limit between sensors, etc.

[0048] Here, the configuration scheme of the sensor is the configuration number and the configuration position of the sensor. Assuming that for sensor type i, the deployment number and the deployment position are the decision variable vectors X of the configuration problem of this type i , the optimization target is f i (x), the multi-task optimization algorithm optimizes the optimization target of the configuration problem of each type of sensor as follows:

[0049] {x1,x2,…,x K-1 ,x K}=argmin{f1(x),f2(x),…,f K-1 (x),f K (x)}

[0050] Step S3, the multi-task optimization algorithm is used to solve the optimization target, and the optimal sensor configuration scheme corresponding to each sensor type is determined.

[0051] The present application decomposes the sensor configuration of the mine digital twin system with a variety of sensor types according to the sensor type and solves it with the help of the multi-task optimization algorithm, improves the configuration effect of the configuration scheme on each type of sensor, and can effectively solve the problem of uneven configuration of different types of sensors and possible redundancy between different sensors.

[0052] In one embodiment, the multi-task optimization algorithm is used to solve the optimization target in step S3 to determine the optimal sensor configuration scheme corresponding to each sensor type, including:

[0053] Step S31, according to the sensor configuration rule, generate N sensor overall configuration schemes as the initial configuration scheme set, i.e. the initial population, each sensor overall configuration scheme includes the sensor configuration scheme corresponding to all sensor types;

[0054] In this step, N can be set to 100, and the size of the initial population is 100; here, according to the difference of the sensor type, the configuration scheme can be randomly generated, or the commonly used historical configuration scheme of the sensor can be installed as the initial configuration scheme.

[0055] Step S32, if the current population is the initial population, for each sensor overall configuration scheme in the initial population, calculate the optimization target value corresponding to each sensor type based on the sensor configuration scheme corresponding to each sensor type, and record the best sensor type in each sensor overall configuration scheme according to the optimization target value; if the current population is not the initial population, calculate the optimization target value corresponding to the best sensor type in each sensor overall configuration scheme recorded in the current population;

[0056] In this step, the sensor type that performs best in each overall sensor configuration scheme is recorded according to the optimization target value. Specifically, the sensor type with the smallest optimization target value in the overall sensor configuration scheme is selected as the sensor type that performs best in the overall sensor configuration scheme.

[0057] Step S33: Calculate the fitness of each sensor configuration scheme based on the optimization target value corresponding to the best performing sensor type in the overall configuration scheme of each sensor in the current population.

[0058] In this step, the fitness of the scheme is f = 1 / rank, where rank is the optimization target value corresponding to the best-performing sensor type in the overall sensor configuration scheme.

[0059] Step S34: Sort the fitness of the schemes from highest to lowest, and select the overall sensor configuration schemes corresponding to the top N fitness schemes as the next population. If the current optimization objective calculation count FE reaches the set number, the process ends. Then, based on the selected top N overall sensor configuration schemes, determine the optimal sensor configuration scheme for each sensor type. If the current optimization objective calculation count FE does not reach the set number, use crossover to generate offspring based on the next population, determine the best-performing sensor type in the offspring, and perform a mutation operation between the offspring and each individual in the next population to obtain the mutated individuals, which will be used as the current population for the next calculation. Return to step S32. In this step, the maximum number of calculations maxFE = 10000 is set; the mutation operation uses a mutation probability of 0.8 for polynomial mutation.

[0060] In one embodiment, generating offspring using a crossover operation on the basis of the next population includes:

[0061] In the next population, randomly select two overall sensor configuration schemes from the N overall sensor configuration schemes, denoted as a and b;

[0062] If the best-performing sensor type is the same in both overall sensor configuration schemes a and b, then a simulated binary crossover method is used to crossover schemes a and b, generating two offspring: overall sensor configuration scheme c and overall sensor configuration scheme d. Otherwise, crossover is performed with a crossover probability of 0.5, generating two offspring: overall sensor configuration scheme c and overall sensor configuration scheme d. The above operations are performed in total. Second-rate.

[0063] In one embodiment, determining the best-performing sensor type in the offspring includes:

[0064] Two child models are sensor overall configuration scheme c and sensor overall configuration scheme d. For sensor overall configuration scheme c and sensor overall configuration scheme d, two random numbers rand1 and rand2 in the range of 0-1 are generated respectively. If rand1 < 0.5, the best performing sensor type in sensor overall configuration scheme c is determined as the best performing sensor type recorded in sensor overall configuration scheme a. If rand1 ≥ 0.5, the best performing sensor type in sensor overall configuration scheme c is determined as the best performing sensor type recorded in sensor overall configuration scheme b. If rand2 < 0.5, the best performing sensor type in sensor overall configuration scheme d is determined as the best performing sensor type recorded in sensor overall configuration scheme a. If rand2 ≥ 0.5, the best performing sensor type in sensor overall configuration scheme d is determined as the best performing sensor type recorded in sensor overall configuration scheme b.

[0065] In one embodiment, determining the optimal sensor configuration scheme for each sensor type based on the selected top N overall sensor configuration schemes includes:

[0066] For each sensor type, the sensor configuration scheme with the optimal objective value among the N overall sensor configuration schemes is selected as the optimal sensor configuration scheme for that sensor type. Here, optimal objective value can be defined as minimum objective value.

[0067] Similar to the inventive concept of the sensor configuration method for a mine digital twin system based on a multi-task optimization algorithm, this application also provides a sensor configuration device for a mine digital twin system based on a multi-task optimization algorithm. Figure 2 A structural block diagram of a sensor configuration device for a mine digital twin system based on a multi-task optimization algorithm according to an embodiment of this application is shown. The device includes:

[0068] The sensor configuration task determination module 210 is used to determine the sensor configuration task, which includes multiple subtasks, each corresponding to a sensor type. Here, the sensor type can be set to 10 to meet the requirements of measuring different parameters and monitoring different operating states. Each subtask is used to determine the number and location of sensors configured within the corresponding sensor type.

[0069] The optimization objective determination module 220 is used to determine the optimization objective for each subtask. The variables in the optimization objective are the sensor configuration schemes corresponding to the sensor type. Here, the sensor configuration scheme includes the number and location of sensors. Let X be the decision variable vector for sensor type i, where the deployment quantity and location are the decision variables for that type of configuration problem. i The optimization objective is f i(x), then the optimization objective of the multi-task optimization algorithm for optimizing the configuration problem of each type of sensor is:

[0070] {x1,x2,…,x K-1 ,x K}=argmin{f1(x),f2(x),…,f K-1 (x),f K (x)}

[0071] The optimization objective solution module 230 is used to solve the optimization objective using a multi-task optimization algorithm to determine the optimal sensor configuration scheme for each sensor type.

[0072] This application decomposes the sensor configuration of a mine digital twin system with diverse sensor types into tasks based on sensor type and solves them using a multi-task optimization algorithm. This improves the configuration effect of the configuration scheme for each type of sensor and can effectively solve the problems of uneven configuration of different types of sensors and potential redundancy between different sensors.

[0073] The sensor configuration device for a mine digital twin system based on a multi-task optimization algorithm in this embodiment has the same further functional implementation as the sensor configuration method in the previous embodiment, and will not be described in detail here.

[0074] The above descriptions are merely various embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A sensor configuration method for a mine digital twin system based on a multi-task optimization algorithm, characterized in that, include: Step S1: Determine the sensor configuration task, which includes multiple sub-tasks, each of which corresponds to a sensor type; Step S2: Determine the optimization objective for each subtask; the variable in the optimization objective is the sensor configuration scheme corresponding to the sensor type; Step S3: Use a multi-task optimization algorithm to solve the optimization objective and determine the optimal sensor configuration scheme for each sensor type. A multi-task optimization algorithm is used to solve the optimization objective, determining the optimal sensor configuration scheme for each sensor type, including: Step S31: Generate N overall sensor configuration schemes according to the sensor configuration rules, as an initial configuration scheme set, i.e., an initial population. Each overall sensor configuration scheme includes sensor configuration schemes corresponding to all sensor types. Step S32: If the current population is the initial population, then for each sensor overall configuration scheme in the initial population, calculate the optimization target value corresponding to each sensor type based on the sensor configuration scheme corresponding to each sensor type, and record the sensor type that performs best in each sensor overall configuration scheme according to the optimization target value; if the current population is not the initial population, calculate the optimization target value corresponding to the sensor type that performs best in each sensor overall configuration scheme recorded in the current population. Step S33: Calculate the fitness of each sensor configuration scheme based on the optimization target value corresponding to the best performing sensor type in each overall sensor configuration scheme in the current population. Step S34: Sort the fitness of the schemes from largest to smallest, and select the overall sensor configuration schemes corresponding to the top N scheme fitness as the next population; if the current optimization target is calculated multiple times... If the set number of iterations is reached, the process ends. Then, based on the selected N overall sensor configuration schemes, the optimal sensor configuration scheme for each sensor type is determined. If the current optimization objective is to calculate the number of iterations... If the set number of iterations is not reached, a crossover operation is used to generate offspring based on the next population, and the sensor type that performs best in the offspring is determined. The offspring is then mutated with each individual in the next population to obtain the mutated individuals, which are used as the current population for the next calculation. The process then returns to step S32.

2. The method as described in claim 1, characterized in that, in, The offspring are generated using a crossover operation based on the next population, including: Two sensor configuration schemes are randomly selected from the N sensor configuration schemes in the next population, and denoted as a and b. If the best-performing sensor type is the same in both overall sensor configuration scheme a and overall sensor configuration scheme b, then the analog binary crossover method is used to cross over overall sensor configuration scheme a and overall sensor configuration scheme b, generating two offspring, namely overall sensor configuration scheme c and overall sensor configuration scheme d. Otherwise, according to... The crossover probabilities are used to generate two offspring: sensor overall configuration scheme c and sensor overall configuration scheme d. The above operations were performed in total. Second-rate.

3. The method as described in claim 2, characterized in that, in, Determining the best-performing sensor type among the offspring includes: The two sub-species are sensor overall configuration scheme c and sensor overall configuration scheme d; for sensor overall configuration scheme c, generate a random number rand1 in the range of 0-1; for sensor overall configuration scheme d, generate a random number rand2 in the range of 0-1; if Then, the sensor type that performs best in the overall sensor configuration scheme c is determined to be the sensor type that performs best in the overall sensor configuration scheme a. Then, the sensor type that performs best in the overall sensor configuration scheme c is determined to be the sensor type that performs best in the overall sensor configuration scheme b; if Then, the sensor type that performs best in the overall sensor configuration scheme d is determined as the sensor type that performs best in the overall sensor configuration scheme a. Then, the sensor type that performs best in the overall sensor configuration scheme d is determined as the sensor type that performs best in the overall sensor configuration scheme b.

4. The method as described in claim 1, characterized in that, in, Based on the selected N overall sensor configuration schemes, determine the optimal sensor configuration scheme for each sensor type, including: For each sensor type, the sensor configuration scheme with the optimal target value corresponding to that sensor type among the N overall sensor configuration schemes is selected as the optimal sensor configuration scheme corresponding to that sensor type.

5. A sensor configuration device for a mine digital twin system based on a multi-task optimization algorithm, characterized in that, include: A sensor configuration task determination module is used to determine a sensor configuration task, wherein the sensor configuration task includes multiple sub-tasks, and each sub-task corresponds to a sensor type. An optimization objective determination module is used to determine the optimization objective for each of the sub-tasks; the variable in the optimization objective is the sensor configuration scheme corresponding to the sensor type; The optimization objective solution module is used to solve the optimization objective using a multi-task optimization algorithm to determine the optimal sensor configuration scheme for each sensor type. The optimization objective solution module is also used to implement the following functions: Step S31: Generate N overall sensor configuration schemes according to the sensor configuration rules, as an initial configuration scheme set, i.e., an initial population. Each overall sensor configuration scheme includes sensor configuration schemes corresponding to all sensor types. Step S32: If the current population is the initial population, then for each sensor overall configuration scheme in the initial population, calculate the optimization target value corresponding to each sensor type based on the sensor configuration scheme corresponding to each sensor type, and record the sensor type that performs best in each sensor overall configuration scheme according to the optimization target value. If the current population is not the initial population, then calculate the optimization target value corresponding to the best performing sensor type in each overall sensor configuration scheme recorded in the current population; Step S33: Calculate the fitness of each sensor configuration scheme based on the optimization target value corresponding to the best performing sensor type in each overall sensor configuration scheme in the current population. Step S34: Sort the fitness of the schemes from largest to smallest, and select the overall sensor configuration schemes corresponding to the top N scheme fitness as the next population; if the current optimization target is calculated multiple times... If the set number of iterations is reached, the process ends. Then, based on the selected N overall sensor configuration schemes, the optimal sensor configuration scheme for each sensor type is determined. If the current optimization objective is to calculate the number of iterations... If the set number of iterations is not reached, a crossover operation is used to generate offspring based on the next population, and the sensor type that performs best in the offspring is determined. The offspring is then mutated with each individual in the next population to obtain the mutated individuals, which are used as the current population for the next calculation. The process then returns to step S32.

6. The apparatus as claimed in claim 5, characterized in that, in, The optimization objective solving module is also used for: Two sensor configuration schemes are randomly selected from the N sensor configuration schemes in the next population, and denoted as a and b. If the best-performing sensor type is the same in both overall sensor configuration scheme a and overall sensor configuration scheme b, then the analog binary crossover method is used to cross over overall sensor configuration scheme a and overall sensor configuration scheme b, generating two offspring, namely overall sensor configuration scheme c and overall sensor configuration scheme d. Otherwise, according to... The crossover probabilities are used to generate two offspring: sensor overall configuration scheme c and sensor overall configuration scheme d. The above operations were performed in total. Second-rate.

7. The apparatus as claimed in claim 6, characterized in that, in, The optimization objective solving module is also used for: The two sub-species are sensor overall configuration scheme c and sensor overall configuration scheme d; for sensor overall configuration scheme c, generate a random number rand1 in the range of 0-1; for sensor overall configuration scheme d, generate a random number rand2 in the range of 0-1; if Then, the sensor type that performs best in the overall sensor configuration scheme c is determined to be the sensor type that performs best in the overall sensor configuration scheme a. Then, the sensor type that performs best in the overall sensor configuration scheme c is determined to be the sensor type that performs best in the overall sensor configuration scheme b; if Then, the sensor type that performs best in the overall sensor configuration scheme d is determined as the sensor type that performs best in the overall sensor configuration scheme a. Then, the sensor type that performs best in the overall sensor configuration scheme d is determined as the sensor type that performs best in the overall sensor configuration scheme b.

8. The apparatus as claimed in claim 5, characterized in that, The optimization objective solving module is also used for: For each sensor type, the sensor configuration scheme with the optimal target value corresponding to that sensor type among the N overall sensor configuration schemes is selected as the optimal sensor configuration scheme corresponding to that sensor type.

Images