Disaster recovery site low-voltage sensor deployment method and system and electronic equipment

Abstract

The invention relates to a method and a system for deploying low-voltage sensors on a disaster recovery site, and electronic equipment. The method comprises the following steps: collecting related data of the low-voltage sensors in the site before a disaster; establishing an observability objective function, an economical efficiency objective function and an efficiency objective function based on the related data; performing weighted calculation on the observability objective function, the economical efficiency objective function and the efficiency objective function to obtain a weighted function; constructing a multi-objective optimization model by taking a minimum weighting function as an objective and taking a deployment space of a disaster recovery site as a constraint; solving the multi-objective optimization model by using an optimization algorithm based on active learning to obtain a multi-objective solution; and deploying the disaster recovery site low-voltage sensor based on the multi-target solution.

Description

Technical Field

[0001] This invention relates to the field of sensor deployment technology, and mainly to a method, system and electronic device for deploying low-voltage sensors in disaster recovery sites. Background Technology

[0002] Low-voltage monitoring at the power system disaster recovery site is a crucial support for rapid power restoration after a disaster and for ensuring accurate monitoring of the system after recovery. As the core component of this monitoring system, the efficiency and rationality of sensor deployment directly determine the overall effectiveness of the power system disaster recovery response. Under normal operating conditions, the distribution of measuring transformers in the target area is clear, and the deployment status can be accurately presented by establishing a sensor distribution point cloud model; however, after a disaster, the complexity of the on-site measurement environment increases sharply, and the original sensor distribution system becomes completely ineffective.

[0003] Currently, there are two major problems in the deployment of low-pressure sensors at disaster recovery sites: First, there is a lack of optimized sensor placement methods adapted to the complex measurement environment of disasters, making it impossible to effectively balance the requirements of system observability and installation economy; second, traditional deployment strategies have not constructed a scientific optimization objective function and constraint mechanism, resulting in disordered steps in the construction of emergency metering devices at disaster sites, which seriously restricts the efficiency of sensor installation and operation; after a disaster, the redeployment of sensors relies on manual experience or simple rules, lacking a systematic optimization process.

[0004] Therefore, there is an urgent need for a sensor deployment optimization method that can solve the problems of low deployment efficiency and unreasonable placement in traditional methods. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention proposes a method, system, and electronic device for deploying low-pressure sensors at disaster recovery sites.

[0006] The technical solution of the present invention is as follows: On one hand, this invention proposes a method for deploying low-pressure sensors at disaster recovery sites, the method comprising: Collect relevant data from low-pressure sensors at the disaster site before the disaster. Based on relevant data, establish observability objective function, economic objective function, and efficiency objective function; calculate the weighted objective function by weighting the observability objective function, economic objective function, and efficiency objective function to obtain the weighted function; A multi-objective optimization model is constructed with the goal of minimizing the weighted function and the deployment space of the disaster recovery site as a constraint. A multi-objective optimization model is solved using an active learning-based optimization algorithm to obtain the multi-objective solution; Deployment of low-pressure sensors at disaster recovery sites based on multi-objective solutions.

[0007] Preferably, an observability objective function is established based on relevant data, specifically as follows: A measurement matrix is ​​constructed based on relevant data, with the following dimensions: ,in This indicates the number of sensors that survived the disaster. This indicates the number of sensors added after the disaster. Indicates the number of related data; The ratios between the standard deviation of the measurement noise and the standard deviation of the measurement signal for each sensor are summed to obtain the summation result. An observability objective function is established based on the rank of the measurement matrix, the cumulative result, and the number of related data.

[0008] Preferably, an economic objective function is established based on relevant data, specifically as follows: Calculation Supplement The difference between the observability value after each sensor and the preset pre-disaster observability value; Based on supplements The economic objective function is established by comparing the total cost after adding each sensor with the difference mentioned above.

[0009] Preferably, an efficiency objective function is established based on relevant data, specifically as follows: Calculation Supplement The deployment time term is obtained by comparing the actual deployment time of each sensor with the theoretical minimum deployment time. Computing Deployment The resource utilization item is obtained by comparing the used resources of each sensor with the total deployed resources. Based on the deployment time and resource utilization terms, the efficiency objective function is obtained.

[0010] Preferably, the multi-objective optimization model is expressed by the following formula: ; ; In the formula, This represents a multi-objective optimization model; Represent the objective function for observability; This represents the economic objective function; Represent the efficiency objective function; This indicates the maximum value of the deployment space constraint; Represents the pre-defined observability objective function weights; This represents the weights of the pre-defined economic objective function; Represents the weights of the preset efficiency objective function; Indicates the deployment space constraint value; Indicates constraints; Indicates the deployment of the first The space occupied by each sensor; Indicates deployable space; Indicates the preset spatial shape factor; Indicates the deployment of the first The distance from each sensor to the obstacle; This indicates the maximum possible distance to obstacles within the deployment area.

[0011] Preferably, the optimization algorithm based on active learning includes an iterative process and an active learning process, wherein the iterative process specifically includes the following steps: Set the feasible region of the solution as , Indicates the lower bound. The upper bound and initial step size are represented; the feasible region includes several randomly generated solutions, each solution including the number of supplementary sensors, the location of the sensors, the space occupied by the sensors, the distance from the sensors to the obstacles, and the deployment resource allocation ratio; Extract from feasible region Initial sample solutions Calculate each Multi-objective value Based on preset filtering criteria, from Select the initial set of non-dominated points And calculate the hypervolume of the initial reference set. ; by As a training set, the Gaussian model corresponding to each objective function is trained to obtain the initial mean and standard deviation for each objective function. The objective functions include observability objective function, economic objective function and efficiency objective function. For the Sub-iteration reference set Update , ; Iterate through the above steps to obtain the set of non-dominated points for each iteration; select the set of non-dominated points with the smallest supervolume from all the non-dominated point sets. ; Will and the Sub-iteration reference set After removing the dominant point, with the first The set of nondominated points in the next iteration Merge to obtain the first Sub-iteration reference set That is, initializing the set of non-dominated temporary points. .

[0012] Preferably, the active learning process includes the following steps: From the initial sample solution Extraction candidate solution set ; based on Each candidate solution Calculate the standard deviation corresponding to the objective function for each candidate solution. Uncertainty; Based on preset weights and each candidate solution The uncertainty for each candidate solution Prioritize the scores and sort them in descending order; Before selection The candidate solutions with the highest scores are merged into... In the process, we obtain the updated temporary set of non-dominated points; After updating the temporary non-dominated point set and selecting the non-dominated point set with the smallest supervolume from all current iterations, merge this set into the updated temporary non-dominated point set to obtain the th... The set of nondominated points in the next iteration ; by As a new training set, the parameters of the Gaussian model corresponding to each objective function are updated to obtain the new mean and new standard deviation for each objective function; calculate and The difference in supervolume is used to obtain the supervolume increment; If the overvolume increment is less than a preset stopping criterion threshold within a preset number of consecutive cycles, then output... If the above conditions are not met, the iteration continues until the non-dominated point set obtained in the last iteration is taken as the final multi-objective solution.

[0013] Preferred options also include: Calculate the descent direction and the optimal value of the descent direction for each nondominated point in the updated temporary nondominated point set; If the optimal value in the direction is less than the preset stability threshold, the initial step size is adjusted based on the uncertainty of the current candidate solution.

[0014] On the other hand, the present invention also provides a deployment system for low-pressure sensors at disaster recovery sites, the system comprising: The data acquisition module collects relevant data from low-pressure sensors at the disaster site before the disaster. The multi-objective solution module establishes observability, economic, and efficiency objective functions based on relevant data; it then calculates a weighted function by weighting the observability, economic, and efficiency objective functions. A multi-objective optimization model is constructed with the goal of minimizing the weighted function and the deployment space of the disaster recovery site as a constraint. A multi-objective optimization model is solved using an active learning-based optimization algorithm to obtain the multi-objective solution; The deployment module deploys low-pressure sensors at disaster recovery sites based on multi-objective solutions.

[0015] In another aspect, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described in the present invention.

[0016] The present invention has the following beneficial effects: 1. This invention provides a method for deploying low-voltage sensors at disaster recovery sites. Based on multi-objective optimization theory, it constructs three objective functions: observability, economy, and efficiency, with "minimization" as the common goal. At the same time, it incorporates deployment space constraints to form a multi-objective optimization model, avoiding the imbalance problem caused by single-objective optimization. In the complex post-disaster environment, it simultaneously meets the requirements of accuracy of low-voltage monitoring of the power system, controllability of deployment costs, and timeliness of emergency response. 2. This invention provides a method for deploying low-pressure sensors at disaster recovery sites. Based on an adaptive step size and active learning algorithm, it improves the efficiency and accuracy of optimization solutions. Through an iterative process, it filters the set of non-dominated points and trains a Gaussian model of the objective function. The active learning process prioritizes candidate solutions with high uncertainty to update the training set. At the same time, it dynamically adjusts the step size according to the optimal value of the direction, balancing the solution speed and the quality of the solution. It solves the problems of traditional optimization algorithms such as "slow convergence, easy getting trapped in local optima, and poor adaptability to complex scenarios", quickly finds the optimal sensor deployment scheme suitable for disaster recovery sites, and adapts to the dynamic changes in the post-disaster environment. It also enhances the adaptability to changes in spatial constraints and differences in obstacle distribution at disaster recovery sites. Attached Figure Description

[0017] Figure 1 This is a detailed flowchart of an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0020] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0021] The terms “comprising” and “including” indicate the presence of the described feature, whole, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0022] The term “and / or” refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes these combinations.

[0023] Example 1: See Figure 1 This invention provides a method for deploying low-pressure sensors at disaster recovery sites, the method comprising: S1. Collect relevant data from low-pressure sensors at the disaster site before the disaster. The relevant data includes the distribution of voltage sensors, current sensors, and their installation locations, quantities, and connection methods. S2. Establish observability objective function, economic objective function and efficiency objective function based on relevant data; S21. The core of electrical safety is achieving system observability standards, that is, by supplementing... A sensor is used to make the rank of the measurement matrix cover all parameters, while suppressing the influence of measurement noise, and an observability objective function is established based on relevant data; A measurement matrix is ​​constructed based on relevant data, with the following dimensions: ,in This indicates the number of sensors that survived the disaster. This indicates the number of sensors added after the disaster. Indicates the number of related data; The ratios between the standard deviation of the measurement noise and the standard deviation of the measurement signal for each sensor are summed to obtain the summation result. Based on the rank of the measurement matrix, the cumulative result, and the number of relevant data points, an observability objective function is established, expressed by the formula: ; In the formula, Let represent the observability objective function, where A smaller value indicates greater observability and higher electrical safety. Indicates supplement Observability values ​​after each sensor; Indicates supplement Measurement matrix after each sensor rank; Indicates the first The standard deviation of sensor measurement noise is used to measure the magnitude of measurement noise. The lower the noise, the more accurate the measurement. Measurement noise refers to the deviation between the collected data and the true value caused by various interference factors (such as environmental interference, electronic noise of the equipment itself, etc.) during the process of sensor data acquisition. Indicates the first The standard deviation of the sensor measurement signal is used to reflect the fluctuation of the measurement signal itself; the measurement signal refers to the voltage and current collected by the voltage and current sensors from the voltage transformer and current transformer. S22. Based on relevant data, establish an economic objective function, specifically as follows: Calculation Supplement The difference between the observability value after each sensor and the preset pre-disaster observability value; Based on supplements The economic objective function is established by comparing the total cost after adding one sensor with the difference mentioned above, and is expressed by the formula: ; In the formula, Let represent the economic objective function, where A smaller value indicates a lower cost per unit increase in observability, resulting in better economic efficiency. Indicates supplement Economic value after one sensor; Indicates supplement The total cost of a sensor, including all expenses such as purchase, installation, and maintenance. Indicates the pre-disaster observability value; S23. Establish an efficiency objective function based on relevant data, specifically as follows: Calculation Supplement The deployment time term is obtained by comparing the actual deployment time of each sensor with the theoretical minimum deployment time. Computing Deployment The resource utilization item is obtained by comparing the used resources of each sensor with the total deployed resources. Based on the deployment time and resource utilization terms, the efficiency objective function is obtained, expressed by the formula: ; In the formula, Let represent the efficiency objective function, where The smaller the value, the closer the actual time and resources match the theoretical optimal value, and the higher the deployment efficiency. Indicates supplement Efficiency values ​​of individual sensors; Indicates supplement The actual deployment time of each sensor; Indicates supplement The theoretical minimum deployment time for a single sensor; Indicates deployment The amount of resources used by each sensor, such as the number of personnel and equipment involved in the deployment; This indicates the total amount of deployed resources, such as the maximum number of personnel, equipment, and other resources that can be mobilized on-site. S3. The observability objective function, economic objective function, and efficiency objective function are weighted to obtain the weighted function; With the goal of minimizing the weighted function and the deployment space of the disaster recovery site as a constraint, a multi-objective optimization model is constructed, expressed by the formula: ; ; In the formula, This represents a multi-objective optimization model; This indicates the maximum value of the deployment space constraint; Represents the pre-defined observability objective function weights; This represents the weights of the pre-defined economic objective function; Represents the weights of the preset efficiency objective function; Indicates the deployment space constraint value; Indicates constraints; Indicates the deployment of the first The space occupied by each sensor; Indicates deployable space; Indicates the preset spatial shape factor; Indicates the deployment of the first The distance from each sensor to the obstacle; Indicates the maximum possible distance to obstacles within the deployment area; S4. Solve the multi-objective optimization model using an active learning-based optimization algorithm to obtain the multi-objective solution; the active learning-based optimization algorithm includes an iterative process and an active learning process; S41. Iterative process, the specific steps are as follows: Set the feasible region of the solution as , Indicates the lower bound. Indicates the upper bound and initial step size. and initialization of non-monotonic memory parameters ; The feasible region is the set of all sensor deployment schemes that meet the on-site constraints of disaster recovery. Essentially, it defines "which deployment schemes are physically and rule-wise feasible", providing a clear range of values ​​for the subsequent extraction of initial sample solutions. The "solution" of the feasible domain is a combination of a set of deployment parameters. The feasible domain includes several randomly generated solutions. Each solution includes the number of supplementary sensors, the location of the sensors, the space occupied by the sensors, the distance from the sensors to the obstacles, and the deployment resource allocation ratio. From the feasible domain Extraction Initial sample solutions Calculate each Multi-objective value Based on preset filtering criteria, from Select the initial set of non-dominated points And calculate the hypervolume of the initial reference set. The preset filtering conditions are: none ; In this embodiment, the hypervolume is an intermediate parameter in the active learning process, providing a... This allows us to obtain a hypervolume, and the algorithm for obtaining this hypervolume can be customized; for example... , Represents a linear function, where... That is ; by As a training set, the Gaussian model corresponding to each objective function is trained to obtain the initial mean and standard deviation for each objective function. The objective functions include observability objective function, economic objective function and efficiency objective function. For the first Sub-iteration reference set Update , ; Iterate through the above steps to obtain the set of nondominated points for each iteration. Select the set of non-dominated points with the smallest supervolume from all non-dominated point sets. ; Will and the Sub-iteration reference set After removing the dominator, we get , With the The set of nondominated points in the next iteration Merge to obtain the first Sub-iteration reference set That is, initializing the set of non-dominated temporary points. ; S42. The active learning process, specifically the following steps: From the initial sample solution Extraction candidate solution set ; based on Each candidate solution Calculate the standard deviation corresponding to the objective function for each candidate solution. Uncertainty ,in Indicates the first The standard deviation of each objective function; Based on preset weights and each candidate solution The uncertainty for each candidate solution Priority scoring is represented as , Indicates the uncertainty weight. Indicates the Pareto correlation weight. This indicates that if the candidate solution If it is tolerance, this item is 1; otherwise, it is 0. The scores are then sorted in descending order. Before selection The candidate solutions with the highest scores are merged into... In the process, the updated temporary non-dominated point set is obtained. , This represents the set of candidate solutions with the highest score; In the updated temporary set of nondominated points and the set of nondominated points obtained from all current iterations Select the set of nondominated points with the smallest supervolume Merge it into the updated temporary non-dominated point set to obtain the first... The set of nondominated points in the next iteration ; by As a new training set, the parameters of the Gaussian model corresponding to each objective function are updated to obtain the new mean and new standard deviation for each objective function; calculate and The difference in supervolume yields the supervolume increment. ; If, within a preset number of consecutive iterations, the increase in supervolume is always less than a preset stopping criterion threshold, then... Then output If the above conditions are not met, the iteration continues until the non-dominated point set obtained in the last iteration is taken as the final multi-objective solution. S43, also includes: calculating the descent direction and the optimal value of the descent direction for each non-dominated point in each updated temporary non-dominated point set, to ensure that subsequent line searches can proceed in the direction of the optimization target while satisfying boundary constraints; The descent direction is expressed by the formula: ; In the formula, Indicates candidate solutions The direction of descent; Indicates candidate solutions In the Gradient of each objective function; Indicates the transpose operation; This represents the preset regularization parameters; in, This indicates a regular expression term, avoiding overly broad directions. If the descent direction is optimal Less than the preset stability threshold Then, based on the uncertainty of the current candidate solution, the initial step size is adjusted, as expressed by the formula: ; In the formula, Indicates the updated step size; This represents the step size adjustment coefficient; The sign function is used to determine whether the step size is increasing or decreasing based on the sign of the uncertainty function. hour, ,when hour, ,when hour, ; The sensitivity coefficient representing the effect of the uncertainty function; express The variance is used to measure the degree of dispersion of uncertainty. The larger the variance, the more drastic the fluctuation of uncertainty. Otherwise, then the candidate solution Approximately Pareto stability, no adjustment to the initial step size; S5. Deploy low-voltage sensors at disaster recovery sites based on multi-objective solutions; The disaster recovery site refers to the actual site used to carry out "post-disaster emergency backup" and "rapid power restoration" after the power system (such as substations, distribution rooms, and low-voltage power supply line areas) encounters natural disasters (earthquakes, floods, typhoons) or human-caused failures (equipment burnout, line breakage). The multi-objective solution is in the form of a matrix, where each matrix value represents the number of low-pressure sensors needed to be added to each disaster recovery site area; for example, the disaster recovery site is divided into... For each region, the multi-objective solution is represented as: ,in Indicates the first Line number List the number of low-pressure sensors needed in each area.

[0024] Example 2: The present invention also provides a deployment system for low-pressure sensors at disaster recovery sites, the system comprising: The data acquisition module collects relevant data from low-pressure sensors at the disaster site before the disaster. The multi-objective solution module establishes observability, economic, and efficiency objective functions based on relevant data; it then calculates a weighted function by weighting the observability, economic, and efficiency objective functions. A multi-objective optimization model is constructed with the goal of minimizing the weighted function and the deployment space of the disaster recovery site as a constraint. A multi-objective optimization model is solved using an active learning-based optimization algorithm to obtain the multi-objective solution; The deployment module deploys low-pressure sensors at disaster recovery sites based on multi-objective solutions.

[0025] T1. Collect relevant data from low-pressure sensors at the disaster site before the disaster. The relevant data includes the distribution of voltage sensors, current sensors, and their installation locations, quantities, and connection methods. T2. Establish observability objective function, economic objective function, and efficiency objective function based on relevant data; T21. The core of electrical safety is achieving system observability standards, that is, by supplementing... A sensor is used to make the rank of the measurement matrix cover all parameters, while suppressing the influence of measurement noise, and an observability objective function is established based on relevant data; A measurement matrix is ​​constructed based on relevant data, with the following dimensions: ,in This indicates the number of sensors that survived the disaster. This indicates the number of sensors added after the disaster. Indicates the number of related data; The ratios between the standard deviation of the measurement noise and the standard deviation of the measurement signal for each sensor are summed to obtain the summation result. Based on the rank of the measurement matrix, the cumulative result, and the number of relevant data points, an observability objective function is established, expressed by the formula: ; In the formula, Let represent the observability objective function, where A smaller value indicates greater observability and higher electrical safety. Indicates supplement Observability values ​​after each sensor; Indicates supplement Measurement matrix after each sensor rank; Indicates the first The standard deviation of sensor measurement noise is used to measure the magnitude of measurement noise. The lower the noise, the more accurate the measurement. Measurement noise refers to the deviation between the collected data and the true value caused by various interference factors (such as environmental interference, electronic noise of the equipment itself, etc.) during the process of sensor data acquisition. Indicates the first The standard deviation of the sensor measurement signal is used to reflect the fluctuation of the measurement signal itself; the measurement signal refers to the voltage and current collected by the voltage and current sensors from the voltage transformer and current transformer. T22. Establish an economic objective function based on relevant data, specifically as follows: Calculation Supplement The difference between the observability value after each sensor and the preset pre-disaster observability value; Based on supplements The economic objective function is established by comparing the total cost after adding one sensor with the difference mentioned above, and is expressed by the formula: ; In the formula, Let represent the economic objective function, where A smaller value indicates a lower cost per unit increase in observability, resulting in better economic efficiency. Indicates supplement Economic value after one sensor; Indicates supplement The total cost of a sensor, including all expenses such as purchase, installation, and maintenance. Indicates the pre-disaster observability value; T23. Establish an efficiency objective function based on relevant data, specifically: Calculation Supplement The deployment time term is obtained by comparing the actual deployment time of each sensor with the theoretical minimum deployment time. Computing Deployment The resource utilization item is obtained by comparing the used resources of each sensor with the total deployed resources. Based on the deployment time and resource utilization terms, the efficiency objective function is obtained, expressed by the formula: ; In the formula, Let represent the efficiency objective function, where The smaller the value, the closer the actual time and resources match the theoretical optimal value, and the higher the deployment efficiency. Indicates supplement Efficiency values ​​of individual sensors; Indicates supplement The actual deployment time of each sensor; Indicates supplement The theoretical minimum deployment time for a single sensor; Indicates deployment The amount of resources used by each sensor, such as the number of personnel and equipment involved in the deployment; This indicates the total amount of deployed resources, such as the maximum number of personnel, equipment, and other resources that can be mobilized on-site. T3. The observability objective function, economic objective function, and efficiency objective function are weighted to obtain the weighted function; With the goal of minimizing the weighted function and the deployment space of the disaster recovery site as a constraint, a multi-objective optimization model is constructed, expressed by the formula: ; ; In the formula, This represents a multi-objective optimization model; This indicates the maximum value of the deployment space constraint; Represents the pre-defined observability objective function weights; This represents the weights of the pre-defined economic objective function; Represents the weights of the preset efficiency objective function; Indicates the deployment space constraint value; Indicates constraints; Indicates the deployment of the first The space occupied by each sensor; Indicates deployable space; Indicates the preset spatial shape factor; Indicates the deployment of the first The distance from each sensor to the obstacle; Indicates the maximum possible distance to obstacles within the deployment area; T4. Solve the multi-objective optimization model using an active learning-based optimization algorithm to obtain the multi-objective solution; the active learning-based optimization algorithm includes an iterative process and an active learning process; T41. Iterative process, the specific steps are as follows: Set the feasible region of the solution as , Indicates the lower bound. Indicates the upper bound and initial step size. and initialization of non-monotonic memory parameters ; The feasible region is the set of all sensor deployment schemes that meet the on-site constraints of disaster recovery. Essentially, it defines "which deployment schemes are physically and rule-wise feasible", providing a clear range of values ​​for the subsequent extraction of initial sample solutions. The "solution" of the feasible domain is a combination of a set of deployment parameters. The feasible domain includes several randomly generated solutions. Each solution includes at least the number of supplementary sensors, the location of the sensors, the space occupied by the sensors, the distance from the sensors to the obstacles, and the deployment resource allocation ratio. From the feasible domain Extraction Initial sample solutions Calculate each Multi-objective value Based on preset filtering criteria, from Select the initial set of non-dominated points And calculate the hypervolume of the initial reference set. The preset filtering conditions are: none ; In this embodiment, the hypervolume is an intermediate parameter in the active learning process, providing a... This allows us to obtain a hypervolume, and the algorithm for obtaining this hypervolume can be customized; for example... , Represents a linear function, where... That is ; by As a training set, the Gaussian model corresponding to each objective function is trained to obtain the initial mean and standard deviation for each objective function. The objective functions include observability objective function, economic objective function and efficiency objective function. For the Sub-iteration reference set Update , ; Iterate through the above steps to obtain the set of nondominated points for each iteration. Select the set of non-dominated points with the smallest supervolume from all non-dominated point sets. ; Will and the Sub-iteration reference set After removing the dominator, we get , With the The set of nondominated points in the next iteration Merge to obtain the first Sub-iteration reference set That is, initializing the set of non-dominated temporary points. ; T42. The active learning process, specifically the following steps: From the initial sample solution Extraction candidate solution set ; based on Each candidate solution Calculate the standard deviation corresponding to the objective function for each candidate solution. Uncertainty ,in Indicates the first The standard deviation of each objective function; Based on preset weights and each candidate solution The uncertainty for each candidate solution Priority scoring is represented as , Indicates the uncertainty weight. Indicates the Pareto correlation weight. This indicates that if the candidate solution If it is tolerance, this item is 1; otherwise, it is 0. The scores are then sorted in descending order. Before selection The candidate solutions with the highest scores are merged into... In the process, the updated temporary non-dominated point set is obtained. , This represents the set of candidate solutions with the highest score; In the updated temporary set of nondominated points and the set of nondominated points obtained from all current iterations Select the set of nondominated points with the smallest supervolume Merge it into the updated temporary non-dominated point set to obtain the first... The set of nondominated points in the next iteration ; by As a new training set, the parameters of the Gaussian model corresponding to each objective function are updated to obtain the new mean and new standard deviation for each objective function; calculate and The difference in supervolume yields the supervolume increment. ; If, within a preset number of consecutive iterations, the increase in supervolume is always less than a preset stopping criterion threshold, then... Then output If the above conditions are not met, the iteration continues until the non-dominated point set obtained in the last iteration is taken as the final multi-objective solution. T43 also includes: calculating the descent direction and the optimal value of the descent direction for each non-dominated point in the updated temporary non-dominated point set, to ensure that the subsequent line search can proceed in the direction of the optimization target while satisfying the boundary constraints; The descent direction is expressed by the formula: ; In the formula, Indicates candidate solutions The direction of descent; Indicates candidate solutions In the Gradient of each objective function; Indicates the transpose operation; This represents the preset regularization parameters; in, This indicates a regular expression term, avoiding overly broad directions. If the descent direction is optimal Less than the preset stability threshold Then, based on the uncertainty of the current candidate solution, the initial step size is adjusted, as expressed by the formula: ; In the formula, Indicates the updated step size; This represents the step size adjustment coefficient; The sign function is used to determine whether the step size is increasing or decreasing based on the sign of the uncertainty function. hour, ,when hour, ,when hour, ; The sensitivity coefficient representing the effect of the uncertainty function; express The variance is used to measure the degree of dispersion of uncertainty. The larger the variance, the more drastic the fluctuation of uncertainty. Otherwise, then the candidate solution Approximately Pareto stability, no adjustment to the initial step size; T5. Deployment of low-pressure sensors at disaster recovery sites based on multi-objective solutions; The disaster recovery site refers to the actual site used to carry out "post-disaster emergency backup" and "rapid power restoration" after the power system (such as substations, distribution rooms, and low-voltage power supply line areas) encounters natural disasters (earthquakes, floods, typhoons) or human-caused failures (equipment burnout, line breakage). The multi-objective solution is in the form of a matrix, where each matrix value represents the number of low-pressure sensors needed to be added to each disaster recovery site area; for example, the disaster recovery site is divided into... For each region, the multi-objective solution is represented as: ,in Indicates the first Line number List the number of low-pressure sensors needed in each area.

[0026] Example 3: This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a method for deploying a low-pressure sensor at a disaster recovery site as described in any one of Embodiment 1.

[0027] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, A and B simultaneously, or B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one of a, b, and c can represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, and c can be single or multiple.

[0028] Those skilled in the art will recognize that the units and algorithm steps described in the embodiments disclosed herein can be implemented using electronic hardware, computer software, or a combination of electronic hardware and software. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0029] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0030] In the several embodiments provided in this application, any function, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0031] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for deploying low-voltage sensors at disaster recovery sites, characterized in that, The method includes: Collect relevant data from low-pressure sensors at the disaster site before the disaster. Based on relevant data, establish observability objective function, economic objective function, and efficiency objective function; calculate the weighted objective function by weighting the observability objective function, economic objective function, and efficiency objective function to obtain the weighted function; A multi-objective optimization model is constructed with the goal of minimizing the weighted function and the deployment space of the disaster recovery site as a constraint. A multi-objective optimization model is solved using an active learning-based optimization algorithm to obtain the multi-objective solution; Deployment of low-pressure sensors at disaster recovery sites based on multi-objective solutions.

2. The method for deploying a low-voltage sensor at a disaster recovery site according to claim 1, characterized in that, An observability objective function is established based on relevant data, specifically as follows: A measurement matrix is ​​constructed based on relevant data, with the following dimensions: ,in This indicates the number of sensors that survived the disaster. This indicates the number of sensors added after the disaster. Indicates the number of related data; The ratios between the standard deviation of the measurement noise and the standard deviation of the measurement signal for each sensor are summed to obtain the summation result. An observability objective function is established based on the rank of the measurement matrix, the cumulative result, and the number of related data.

3. The method for deploying a low-voltage sensor at a disaster recovery site according to claim 2, characterized in that, An economic objective function is established based on relevant data, specifically as follows: Calculation Supplement The difference between the observability value after each sensor and the preset pre-disaster observability value; Based on supplements The economic objective function is established by comparing the total cost after adding each sensor with the difference mentioned above.

4. The method for deploying a low-voltage sensor at a disaster recovery site according to claim 2, characterized in that, An efficiency objective function is established based on relevant data, specifically as follows: Calculation Supplement The deployment time term is obtained by comparing the actual deployment time of each sensor with the theoretical minimum deployment time. Computing Deployment The resource utilization item is obtained by comparing the used resources of each sensor with the total deployed resources. Based on the deployment time and resource utilization terms, the efficiency objective function is obtained.

5. The method for deploying a low-voltage sensor at a disaster recovery site according to claim 1, characterized in that, The multi-objective optimization model is expressed by the following formula: ； ； In the formula, This represents a multi-objective optimization model; Represent the objective function for observability; This represents the economic objective function; Represent the efficiency objective function; This indicates the maximum value of the deployment space constraint; Represents the pre-defined observability objective function weights; This represents the weights of the pre-defined economic objective function; Represents the weights of the preset efficiency objective function; Indicates the deployment space constraint value; Indicates constraints; Indicates the deployment of the first The space occupied by each sensor; Indicates deployable space; Indicates the preset spatial shape factor; Indicates the deployment of the first The distance from each sensor to the obstacle; This indicates the maximum possible distance to obstacles within the deployment area.

6. The method for deploying a low-voltage sensor at a disaster recovery site according to claim 1, characterized in that, The optimization algorithm based on active learning includes an iterative process and an active learning process, wherein the iterative process comprises the following steps: Set the feasible region of the solution as , Indicates the lower bound. The upper bound and initial step size are represented; the feasible region includes several randomly generated solutions, each solution including the number of supplementary sensors, the location of the sensors, the space occupied by the sensors, the distance from the sensors to the obstacles, and the deployment resource allocation ratio; Extract from feasible region Initial sample solutions Calculate each Multi-objective value Based on preset filtering criteria, from Select the initial set of non-dominated points And calculate the hypervolume of the initial reference set. ; by As a training set, the Gaussian model corresponding to each objective function is trained to obtain the initial standard deviation corresponding to each objective function, where the objective functions include the observability objective function, the economy objective function, and the efficiency objective function; For the first Sub-iteration reference set Update , ; Iterate through the above steps to obtain the set of nondominated points for each iteration; Select the set of non-dominated points with the smallest supervolume from all non-dominated point sets. ; Will and the Sub-iteration reference set After removing the dominant point, with the first The set of nondominated points in the next iteration Merge to obtain the first Sub-iteration reference set That is, initializing the set of non-dominated temporary points. .

7. The method for deploying a low-voltage sensor at a disaster recovery site according to claim 6, characterized in that, The active learning process consists of the following steps: From the initial sample solution Extraction candidate solution set ; based on Each candidate solution Calculate the standard deviation corresponding to the objective function for each candidate solution. Uncertainty; Based on preset weights and each candidate solution The uncertainty for each candidate solution Prioritize the scores and sort them in descending order; Before selection The candidate solutions with the highest scores are merged into... In the process, we obtain the updated temporary set of non-dominated points; After updating the temporary non-dominated point set and selecting the non-dominated point set with the smallest supervolume from all current iterations, merge this set into the updated temporary non-dominated point set to obtain the th... The set of nondominated points in the next iteration ; by As a new training set, the parameters of the Gaussian model corresponding to each objective function are updated to obtain the new standard deviation for each objective function; calculate and The difference in supervolume is used to obtain the supervolume increment; If the overvolume increment is less than a preset stopping criterion threshold within a preset number of consecutive cycles, then output... If the above conditions are not met, the iteration continues until the non-dominated point set obtained in the last iteration is taken as the final multi-objective solution.

8. The method for deploying a low-voltage sensor at a disaster recovery site according to claim 7, characterized in that, Also includes: Calculate the descent direction and the optimal value of the descent direction for each nondominated point in the updated temporary nondominated point set; If the optimal value in the descent direction is less than the preset stability threshold, the initial step size is adjusted based on the uncertainty of the current candidate solution.

9. A deployment system for low-voltage sensors at disaster recovery sites, characterized in that, The system includes: The data acquisition module collects relevant data from low-pressure sensors at the disaster site before the disaster. The multi-objective solution module establishes observability, economic, and efficiency objective functions based on relevant data; it then calculates a weighted function by weighting the observability, economic, and efficiency objective functions. A multi-objective optimization model is constructed with the goal of minimizing the weighted function and the deployment space of the disaster recovery site as a constraint. A multi-objective optimization model is solved using an active learning-based optimization algorithm to obtain the multi-objective solution; The deployment module deploys low-pressure sensors at disaster recovery sites based on multi-objective solutions.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 8.

Images