Intelligent scheduling method and platform for operation and maintenance equipment of photovoltaic power station

Abstract

The invention discloses an intelligent scheduling method and platform for operation and maintenance equipment of a photovoltaic power station, and relates to the technical field of equipment scheduling. The method comprises the following steps: acquiring a plurality of operation and maintenance work orders of a target photovoltaic power station; traversing the plurality of operation and maintenance work orders, extracting operation and maintenance vectors, and obtaining a plurality of operation and maintenance feature vector sets; performing execution reliability identification on the plurality of operation and maintenance work orders, and determining a plurality of operation and maintenance work order reliability factors; the condition that the scheduling scheme meets operation and maintenance equipment resources is taken as a primary scheduling constraint; determining a plurality of initial device scheduling schemes; performing constraint optimization on the plurality of initial equipment scheduling schemes according to the primary scheduling constraint to obtain a target intelligent scheduling scheme; and calling operation and maintenance equipment resources of the target photovoltaic power station according to the target intelligent scheduling scheme to complete processing of the plurality of operation and maintenance work orders. The technical problems of low scheduling efficiency and unreasonable resource allocation of the photovoltaic power station operation and maintenance equipment in the prior art are solved, and the technical effects of improving the operation and maintenance scheduling efficiency and the equipment resource utilization rate are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment scheduling, and particularly to an intelligent scheduling method and platform for photovoltaic power station operation and maintenance equipment. Background Art

[0002] With the rapid development of photovoltaic power generation technology, the scale of photovoltaic power stations has been continuously expanding, and the complexity of operation and maintenance management has also increased accordingly. Traditional operation and maintenance management of photovoltaic power stations rely on manual scheduling, which has problems such as uneven task allocation, waste of equipment resources, and low work order processing efficiency. Especially during the operation and maintenance of photovoltaic power stations, there are various types of equipment, complex and periodic operation and maintenance tasks. How to efficiently and reasonably schedule operation and maintenance equipment to ensure the timely processing of each work order and the rational use of equipment resources has become the key to improving the operation efficiency of photovoltaic power stations. Current scheduling methods usually fail to fully consider the diversity of operation and maintenance tasks and the constraints of equipment resources, resulting in situations of equipment idleness or overuse. On the other hand, existing technologies lack intelligent scheduling means and cannot dynamically adjust resource allocation according to actual operation and maintenance requirements. Summary of the Invention

[0003] The present application provides an intelligent scheduling method and platform for photovoltaic power station operation and maintenance equipment, which solves the technical problems of low scheduling efficiency and unreasonable resource allocation of photovoltaic power station operation and maintenance equipment in the prior art.

[0004] In view of the above problems, the embodiments of the present application provide an intelligent scheduling method and platform for photovoltaic power station operation and maintenance equipment.

[0005] In the first aspect of the present application, there is provided an intelligent scheduling method for photovoltaic power station operation and maintenance equipment, the method comprising:

[0006] Obtain multiple operation and maintenance work orders of a target photovoltaic power station, wherein each operation and maintenance work order includes a matching operation and maintenance equipment type and a matching operation and maintenance equipment working duration; traverse the multiple operation and maintenance work orders, extract operation and maintenance vectors, and obtain multiple operation and maintenance feature vector sets; based on the multiple operation and maintenance feature vector sets, perform execution reliability identification on the multiple operation and maintenance work orders to determine multiple operation and maintenance work order reliability factors; obtain the operation and maintenance equipment resources of the target photovoltaic power station, and take that the scheduling scheme meets the operation and maintenance equipment resources as the primary scheduling constraint; based on the magnitudes of the multiple operation and maintenance work order reliability factors, combine the matching operation and maintenance equipment type and the matching operation and maintenance equipment working duration corresponding to each operation and maintenance work order in the multiple operation and maintenance work orders to perform operation and maintenance work order processing scheme identification, and determine multiple initial equipment scheduling schemes; perform constraint optimization on the multiple initial equipment scheduling schemes according to the primary scheduling constraint to obtain a target intelligent scheduling scheme; and call the operation and maintenance equipment resources of the target photovoltaic power station according to the target intelligent scheduling scheme to complete the processing of the multiple operation and maintenance work orders.

[0007] The second aspect of the present application provides an intelligent scheduling platform for photovoltaic power station operation and maintenance equipment, and the platform includes:

[0008] An operation and maintenance work order acquisition module: acquiring multiple operation and maintenance work orders of a target photovoltaic power station, where each operation and maintenance work order includes a matching operation and maintenance equipment type and a matching operation and maintenance equipment working duration; an operation and maintenance vector extraction module: traversing the multiple operation and maintenance work orders, extracting operation and maintenance vectors, and obtaining multiple operation and maintenance feature vector sets; a reliability identification module: performing execution reliability identification on the multiple operation and maintenance work orders based on the multiple operation and maintenance feature vector sets, and determining multiple operation and maintenance work order reliability factors; a constraint setting module: acquiring the operation and maintenance equipment resources of the target photovoltaic power station, and taking that the scheduling scheme meets the operation and maintenance equipment resources as the primary scheduling constraint; a scheme identification module: identifying an operation and maintenance work order processing scheme based on the magnitudes of the multiple operation and maintenance work order reliability factors, and in combination with the matching operation and maintenance equipment type and the matching operation and maintenance equipment working duration corresponding to each operation and maintenance work order in the multiple operation and maintenance work orders, and determining multiple initial equipment scheduling schemes; an optimization module: performing constraint optimization on the multiple initial equipment scheduling schemes according to the primary scheduling constraint to obtain a target intelligent scheduling scheme; a processing module: invoking the operation and maintenance equipment resources of the target photovoltaic power station according to the target intelligent scheduling scheme to complete the processing of the multiple operation and maintenance work orders.

[0009] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0010] First, acquire multiple operation and maintenance work orders of a target photovoltaic power station, where each operation and maintenance work order includes a matching operation and maintenance equipment type and a matching operation and maintenance equipment working duration. Next, traverse the multiple operation and maintenance work orders, extract operation and maintenance vectors, and obtain multiple operation and maintenance feature vector sets. Further, perform execution reliability identification on the multiple operation and maintenance work orders based on the multiple operation and maintenance feature vector sets, and determine multiple operation and maintenance work order reliability factors. At the same time, acquire the operation and maintenance equipment resources of the target photovoltaic power station, and take that the scheduling scheme meets the operation and maintenance equipment resources as the primary scheduling constraint. Next, identify an operation and maintenance work order processing scheme based on the magnitudes of the multiple operation and maintenance work order reliability factors, and in combination with the matching operation and maintenance equipment type and the matching operation and maintenance equipment working duration corresponding to each operation and maintenance work order in the multiple operation and maintenance work orders, and determine multiple initial equipment scheduling schemes. Then, perform constraint optimization on the multiple initial equipment scheduling schemes according to the primary scheduling constraint to obtain a target intelligent scheduling scheme. Finally, invoke the operation and maintenance equipment resources of the target photovoltaic power station according to the target intelligent scheduling scheme to complete the processing of the multiple operation and maintenance work orders. This solves the technical problems of low scheduling efficiency of photovoltaic power station operation and maintenance equipment and unreasonable resource allocation in the prior art, and achieves the technical effect of improving operation and maintenance scheduling efficiency and equipment resource utilization rate. Description of the Drawings

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the accompanying drawings required for the description of the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0012] Figure 1 Schematic flow chart of the intelligent scheduling method for photovoltaic power station operation and maintenance equipment provided by the embodiments of the present application;

[0013] Figure 2 Schematic structural diagram of the intelligent scheduling platform for photovoltaic power station operation and maintenance equipment provided by the embodiments of the present application.

[0014] Explanation of reference numerals: Operation and maintenance work order acquisition module 11, operation and maintenance vector extraction module 12, reliability identification module 13, constraint setting module 14, solution identification module 15, optimization module 16, processing module 17. Specific embodiments

[0015] The embodiments of the present application provide an intelligent scheduling method and platform for photovoltaic power station operation and maintenance equipment, which solve the technical problems of low scheduling efficiency and unreasonable resource allocation of photovoltaic power station operation and maintenance equipment in the prior art.

[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application.

[0017] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units does not necessarily have to be limited to those clearly listed steps or units, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products, or devices.

[0018] Embodiment 1, as Figure 1 shown, the embodiments of the present application provide an intelligent scheduling method for photovoltaic power station operation and maintenance equipment, wherein the method includes:

[0019] Obtain multiple operation and maintenance work orders of the target photovoltaic power station, where each operation and maintenance work order includes a matching operation and maintenance equipment type and a matching operation and maintenance equipment working duration.

[0020] Obtain multiple operation and maintenance work orders from the target photovoltaic power station. Each operation and maintenance work order details the operation and maintenance tasks to be performed, including matching the types of operation and maintenance equipment (such as inverters, cleaning robots, inspection drones, etc.) and the working hours of the operation and maintenance equipment (i.e., the time required to complete the task).

[0021] Traverse the multiple operation and maintenance work orders, extract operation and maintenance vectors, and obtain multiple sets of operation and maintenance feature vectors.

[0022] Traverse the obtained multiple operation and maintenance work orders, extract the key information of each work order to form an operation and maintenance feature vector, and combine the operation and maintenance vectors extracted from each operation and maintenance work order to form a set of operation and maintenance feature vectors.

[0023] Furthermore, obtain multiple sets of operation and maintenance feature vectors, including:

[0024] Obtain multiple sample operation and maintenance work orders and multiple sets of sample operation and maintenance feature vectors as a sample data set; perform supervised training on the network layer constructed based on the feedforward neural network according to the sample data set to learn the mapping relationship between the operation and maintenance work order and the operation and maintenance feature vector until the training converges, and obtain the trained vector extraction network layer; use the vector extraction network layer to perform feature vector recognition on the multiple operation and maintenance work orders to obtain the multiple sets of operation and maintenance feature vectors.

[0025] Specifically, collect multiple sample operation and maintenance work orders. Each sample work order contains its corresponding operation and maintenance feature vectors, which can be manually labeled or extracted from historical data and cover various key attributes of the work order, such as equipment type, working hours, task priority, etc.; combine the multiple sample operation and maintenance work orders and the corresponding sets of sample operation and maintenance feature vectors to form a sample data set; according to the obtained sample data set, construct a network layer using a feedforward neural network. This network layer is designed to receive the operation and maintenance work order as input and output the corresponding operation and maintenance feature vector; perform supervised training on the network layer using the sample data set. During the training process, the network layer will try to learn the mapping relationship between the operation and maintenance work order and the operation and maintenance feature vector. By continuously adjusting the weight and bias parameters of the network layer, the output of the network layer gradually approaches the set of sample operation and maintenance feature vectors. When the training error of the network layer reaches a preset threshold or the number of training rounds reaches a preset upper limit, it is considered that the network layer has been trained to converge; after training is completed, obtain a vector extraction network layer that can accurately extract operation and maintenance feature vectors. The vector extraction network layer can be regarded as a mapping function from the operation and maintenance work order to the operation and maintenance feature vector; use the already trained vector extraction network layer to perform feature vector recognition on the multiple operation and maintenance work orders. The network layer will automatically extract the operation and maintenance feature vectors corresponding to the operation and maintenance work orders according to the learned mapping relationship.

[0026] Based on the multiple operation and maintenance feature vector sets, perform execution reliability identification on the multiple operation and maintenance work orders to determine multiple operation and maintenance work order reliability factors.

[0027] Through the previously extracted multiple operation and maintenance feature vector sets, evaluate the execution reliability of each operation and maintenance work order, and determine the operation and maintenance work order reliability factor of each operation and maintenance work order. The operation and maintenance work order reliability factor indicates the degree of credibility that the work order can be completed as expected.

[0028] Furthermore, determining multiple operation and maintenance work order reliability factors includes:

[0029] Retrieve the historical operation and maintenance database, and randomly extract the first operation and maintenance feature vector set from the multiple operation and maintenance feature vector sets. Among them, the historical operation and maintenance database includes multiple historical operation and maintenance feature vector sets of multiple historical operation and maintenance work orders and multiple historical operation and maintenance work order reliability factors; respectively, using any one operation and maintenance feature vector in the first operation and maintenance feature vector set as an index, retrieve the multiple historical operation and maintenance feature vector sets of the multiple historical operation and maintenance work orders in the historical operation and maintenance database to obtain the first reliability identification sub-database. Among them, the first reliability identification sub-database includes multiple first retrieved historical operation and maintenance work order reliability factors; perform trace-intensive analysis on the multiple first retrieved historical operation and maintenance work order reliability factors, and determine the first operation and maintenance work order reliability factor according to the analysis results; perform retrieval and trace-intensive analysis on the multiple operation and maintenance feature vector sets in the historical operation and maintenance database respectively to determine the multiple operation and maintenance work order reliability factors.

[0030] Specifically, data is retrieved from the historical operation and maintenance database, which contains multiple sets of historical operation and maintenance feature vectors of multiple historical operation and maintenance work orders and the corresponding multiple reliable factors of the historical operation and maintenance work orders. Among them, the reliable factor of the historical operation and maintenance work order can be obtained by dividing the difference between the actual operation and maintenance time and the estimated operation and maintenance time by the estimated operation and maintenance time; randomly extract the first set of operation and maintenance feature vectors from the multiple sets of operation and maintenance feature vectors, and use any one of the operation and maintenance feature vectors in the first set of operation and maintenance feature vectors as an index to retrieve the set of historical operation and maintenance feature vectors similar to it in the historical operation and maintenance database. The result of the retrieval is one or more historical operation and maintenance work orders, and the feature vectors of these work orders are the closest to the feature vectors of the current operation and maintenance work order under a certain metric standard (such as Euclidean distance, cosine similarity, etc.); form the first reliability identification sub-database with the retrieved historical operation and maintenance work orders and their corresponding reliable factors; conduct a tracking density analysis on the multiple reliable factors of the first retrieved historical operation and maintenance work orders in the first reliability identification sub-database to identify the successful execution of the historical work orders similar to the current operation and maintenance work order and find out the key factors affecting reliability; according to the results of the tracking density analysis, determine the reliable factor of the first operation and maintenance work order, and the reliable factor of the first operation and maintenance work order represents the possibility of successful execution of the work order; for other operation and maintenance work orders to be evaluated, by repeating the above retrieval and analysis process, finally determine the reliable factor of each operation and maintenance work order.

[0031] Furthermore, conducting a tracking density analysis on the multiple reliable factors of the first retrieved historical operation and maintenance work orders and determining the reliable factor of the first operation and maintenance work order according to the analysis results includes:

[0032] Randomly extract a reliable factor of the first retrieved historical operation and maintenance work order from the multiple reliable factors of the first retrieved historical operation and maintenance work orders as the starting center; construct the starting neighborhood of the starting center based on the preset tracking bandwidth; use the weighted center iteration function to analyze and iterate the multiple reliable factors of the first retrieved historical operation and maintenance work orders and the starting center in the starting neighborhood to obtain the iterative weighted center, and then construct the neighborhood based on the iterative weighted center and conduct iteration again until the preset iteration stop condition is met, that is, the distance between the iterative weighted centers obtained in two adjacent iterations is less than or equal to the preset distance threshold; use the reliable factor of the first retrieved historical operation and maintenance work order corresponding to the target weighted center as the reliable factor of the first operation and maintenance work order.

[0033] Specifically, randomly select one from multiple first retrieval history operation and maintenance work order reliability factors as the starting center, which represents the initial estimate of the current iteration process; based on a preset tracking bandwidth (a distance threshold preset by those skilled in the art), construct the starting neighborhood of the starting center, and the starting neighborhood contains the first retrieval history operation and maintenance work order reliability factors that are closest to the starting center under a certain metric standard (such as Euclidean distance); use the weighted center iteration function to analyze and iterate on multiple first retrieval history operation and maintenance work order reliability factors and the starting center in the starting neighborhood to obtain an iterative weighted center; based on the new iterative weighted center, reconstruct the neighborhood and iterate again; repeat until the preset iteration stop condition is met. The preset iteration stop condition is usually defined as the distance between the iterative weighted centers obtained in two adjacent iterations being less than or equal to the preset distance threshold; when the iteration process meets the stop condition, the obtained target weighted center is the final estimate of the iteration process, and the first retrieval history operation and maintenance work order reliability factor corresponding to the target weighted center is used as the reliability factor of the first operation and maintenance work order.

[0034] Furthermore, construct the weighted center iteration function, where the weighted center iteration function is:

[0035] where Point(x) is the iterative weighted center, N(x) is multiple first retrieval history operation and maintenance work order reliability factors in the starting neighborhood, x i is the i-th first retrieval history operation and maintenance work order reliability factor among multiple first retrieval history operation and maintenance work order reliability factors, x is the starting center, and K(x i -x) is a weight kernel function constructed based on the Gaussian function.

[0036] The weighted center iteration function updates the iterative center in a weighted average manner, adjusts the current center according to the reliability factors of historical operation and maintenance work orders in the neighborhood in each iteration until convergence, and finally obtains a stable target weighted center. Point(x) represents the new weighted center calculated in the iteration process, N(x) represents the set of multiple first retrieval history operation and maintenance work order reliability factors in the starting neighborhood centered on x, x i represents the i-th first retrieval history operation and maintenance work order reliability factor in the set N(x), x represents the current starting center or the weighted center obtained in the previous iteration, and K(x i -x) represents a weight kernel function constructed based on the Gaussian function, which is used to measure the similarity between x i and x.

[0037] The weight kernel function is expressed as where σ is a preset bandwidth parameter that controls the weight distribution between points in the neighborhood and the starting center.

[0038] Obtain the operation and maintenance equipment resources of the target photovoltaic power station, and take that the scheduling plan meets the operation and maintenance equipment resources as the primary scheduling constraint.

[0039] Collect the operation and maintenance equipment resource information of the target photovoltaic power station, including but not limited to equipment type and quantity, equipment performance parameters, equipment status, equipment distribution, etc.; formulate scheduling constraint conditions according to the operation and maintenance equipment resource information. The primary scheduling constraint should ensure that the scheduling plan can make full use of the existing equipment resources and avoid resource idleness or overuse.

[0040] Based on the magnitudes of the reliability factors of the multiple operation and maintenance work orders, in combination with the matching operation and maintenance equipment types and the matching operation and maintenance equipment working hours corresponding to each operation and maintenance work order among the multiple operation and maintenance work orders, identify the operation and maintenance work order processing solutions, and determine multiple initial equipment scheduling plans.

[0041] Based on the magnitudes of the reliability factors of the multiple operation and maintenance work orders, in combination with the matching equipment types and equipment working hours corresponding to each operation and maintenance work order, identify the operation and maintenance work order processing solutions, so as to determine multiple initial equipment scheduling plans. Specifically, sort the work orders according to the priority of the reliability factors; in combination with the matching operation and maintenance equipment types and equipment working hours required by each work order, screen out the appropriate equipment resources for allocation; for the work orders with higher reliability factors, they will be scheduled first to ensure that high-reliability tasks can be executed in a timely manner; for the work orders with longer equipment working hours, ensure that the available time of the equipment meets the task requirements; on this basis, by comprehensively considering the equipment type, working hours and reliability factors, identify the processing solutions of the multiple operation and maintenance work orders, and form multiple initial equipment scheduling plans.

[0042] Conduct constraint optimization on the multiple initial equipment scheduling plans according to the primary scheduling constraint to obtain the target intelligent scheduling plan.

[0043] After obtaining the multiple initial equipment scheduling plans, it is necessary to conduct constraint optimization on these plans according to the primary scheduling constraint determined previously (i.e., the actual situation of the operation and maintenance equipment resources) to obtain the target intelligent scheduling plan.

[0044] Furthermore, obtaining the target intelligent scheduling plan includes:

[0045] Using the primary scheduling constraint to discriminate among the multiple initial device scheduling plans, eliminating the plans that do not meet the primary scheduling constraint, and obtaining multiple screened device scheduling plans; traversing the multiple screened device scheduling plans to perform scheduling fitness analysis, obtaining multiple screened scheduling fitness values; selecting the screened device scheduling corresponding to the maximum value among the multiple screened scheduling fitness values as the optimization target, and adjusting the remaining multiple screened device scheduling plans according to a preset adjustment plan to obtain multiple adjusted device scheduling plans, where the preset adjustment plan is to randomly adjust the scheduling order in the multiple screened device scheduling plans with the scheduling order in the optimization target as the goal; updating the optimization target according to the multiple adjusted device scheduling plans, and after multiple update iterations, taking the adjusted device scheduling plan corresponding to the optimization target obtained in the last iteration as the target intelligent scheduling plan.

[0046] Specifically, using the primary scheduling constraint to discriminate among the multiple initial device scheduling plans, eliminating the scheduling plans that do not meet the device resource constraints (such as device availability, working hours, etc.), so as to obtain multiple screened device scheduling plans; traversing the multiple screened device scheduling plans to perform scheduling fitness analysis, calculating the screened scheduling fitness by analyzing the execution effect of each plan in the actual operation and maintenance tasks (such as the effective utilization of device resources, task completion time, etc.), and a plan with a high fitness indicates that it can better meet the scheduling goal; selecting the screened device scheduling plan corresponding to the maximum value among the multiple screened scheduling fitness values as the optimization target; for the remaining screened device scheduling plans, they will be adjusted according to a preset adjustment plan, and the preset adjustment plan is to randomly adjust the scheduling order in the multiple screened device scheduling plans to find a possibly better plan; updating the optimization target according to the multiple adjusted device scheduling plans, and gradually improving the scheduling plan and enhancing the overall scheduling effect through continuous iteration and optimization; after each iteration, the optimization target will be updated according to the adjusted scheduling plan until the final scheduling plan reaches the optimal; after multiple iteration updates, the adjusted device scheduling plan corresponding to the finally obtained optimization target will be used as the target intelligent scheduling plan.

[0047] Furthermore, it includes:

[0048] Traverse the multiple adjusted device scheduling schemes for scheduling fitness analysis to obtain multiple adjusted scheduling fitnesses; determine whether there is a scheduling fitness greater than or equal to the screening scheduling fitness corresponding to the optimization target among the multiple adjusted scheduling fitnesses. If so, use the adjusted device scheduling scheme corresponding to the maximum value among the multiple adjusted scheduling fitnesses as the new optimization target and continue with the update iteration; if not, add the adjustment method of the obtained multiple adjusted device scheduling schemes to the tabu adjustment method, which is not allowed to be used again within the preset tabu iteration times, and based on the optimization target, adjust the remaining multiple screened device scheduling schemes according to the preset adjustment scheme except the tabu adjustment method, and update and iterate the optimization target according to the adjustment result.

[0049] Specifically, traverse the multiple adjusted device scheduling schemes for scheduling fitness analysis, and calculate the adjusted scheduling fitness of each adjustment scheme; determine whether there is a scheduling fitness greater than or equal to the screening scheduling fitness corresponding to the optimization target among the multiple adjusted scheduling fitnesses; if it exists, it means that the current adjustment scheme may improve the scheduling effect. Therefore, use the adjusted device scheduling scheme corresponding to the maximum value among the multiple adjusted scheduling fitnesses as the new optimization target and continue with the update iteration to obtain a better device scheduling scheme; if there is no better scheme, that is, the current adjustment scheme cannot exceed the existing optimization target, then add the adjustment method of the obtained adjusted device scheduling scheme to the tabu adjustment method. The tabu adjustment method means that this adjustment method is not allowed to be used again within the next preset tabu iteration times to avoid falling into the cycle of local optimal solutions. At the same time, based on the optimization target, adjust the remaining screened device scheduling schemes again, but at this time, the tabu adjustment method will be excluded, and optimization will still be carried out according to other preset adjustment schemes; through this tabu search strategy, it is possible to effectively avoid repeatedly using ineffective or inefficient adjustment methods and continue to update and iterate the remaining schemes according to the optimization target to gradually find the optimal intelligent scheduling scheme. After multiple iteration cycles, an optimal device scheduling scheme that meets the scheduling constraints, optimization target, and task requirements will finally be obtained as the target intelligent scheduling scheme to ensure that the operation and maintenance tasks of the photovoltaic power station can be completed efficiently and accurately.

[0050] Invoke the operation and maintenance device resources of the target photovoltaic power station according to the target intelligent scheduling scheme to complete the processing of the multiple operation and maintenance work orders.

[0051] According to the target intelligent scheduling scheme, allocate and call the eligible operation and maintenance device resources; the scheduled devices will start to execute the corresponding operation and maintenance work order processing tasks according to the target intelligent scheduling scheme.

[0052] In summary, the embodiments of the present application at least have the following technical effects:

[0053] First, obtain multiple operation and maintenance work orders of the target photovoltaic power station. Each operation and maintenance work order includes a matching operation and maintenance equipment type and a matching operation and maintenance equipment working duration. Next, traverse the multiple operation and maintenance work orders, extract operation and maintenance vectors, and obtain multiple operation and maintenance feature vector sets. Further, perform execution reliability identification on the multiple operation and maintenance work orders based on the multiple operation and maintenance feature vector sets to determine multiple operation and maintenance work order reliability factors. At the same time, obtain the operation and maintenance equipment resources of the target photovoltaic power station, and take the scheduling plan meeting the operation and maintenance equipment resources as the primary scheduling constraint. Next, based on the magnitudes of the multiple operation and maintenance work order reliability factors, combine the matching operation and maintenance equipment type and the matching operation and maintenance equipment working duration corresponding to each operation and maintenance work order in the multiple operation and maintenance work orders to identify an operation and maintenance work order processing plan, and determine multiple initial equipment scheduling plans. Then, perform constraint optimization on the multiple initial equipment scheduling plans according to the primary scheduling constraint to obtain a target intelligent scheduling plan. Finally, call the operation and maintenance equipment resources of the target photovoltaic power station according to the target intelligent scheduling plan to complete the processing of the multiple operation and maintenance work orders. This solves the technical problems of low operation and maintenance equipment scheduling efficiency and unreasonable resource allocation in the prior art, and achieves the technical effect of improving operation and maintenance scheduling efficiency and equipment resource utilization rate.

[0054] Embodiment 2, based on the same inventive concept as the intelligent scheduling method for photovoltaic power station operation and maintenance equipment in the foregoing embodiment, as Figure 2 shown, the present application provides an intelligent scheduling platform for photovoltaic power station operation and maintenance equipment. Among them, the platform includes:

[0055] An operation and maintenance work order acquisition module 11: Obtain multiple operation and maintenance work orders of the target photovoltaic power station. Each operation and maintenance work order includes a matching operation and maintenance equipment type and a matching operation and maintenance equipment working duration; An operation and maintenance vector extraction module 12: Traverse the multiple operation and maintenance work orders, extract operation and maintenance vectors, and obtain multiple operation and maintenance feature vector sets; A reliability identification module 13: Perform execution reliability identification on the multiple operation and maintenance work orders based on the multiple operation and maintenance feature vector sets to determine multiple operation and maintenance work order reliability factors; A constraint setting module 14: Obtain the operation and maintenance equipment resources of the target photovoltaic power station, and take the scheduling plan meeting the operation and maintenance equipment resources as the primary scheduling constraint; A plan identification module 15: Based on the magnitudes of the multiple operation and maintenance work order reliability factors, combine the matching operation and maintenance equipment type and the matching operation and maintenance equipment working duration corresponding to each operation and maintenance work order in the multiple operation and maintenance work orders to identify an operation and maintenance work order processing plan, and determine multiple initial equipment scheduling plans; An optimization module 16: Perform constraint optimization on the multiple initial equipment scheduling plans according to the primary scheduling constraint to obtain a target intelligent scheduling plan; A processing module 17: Call the operation and maintenance equipment resources of the target photovoltaic power station according to the target intelligent scheduling plan to complete the processing of the multiple operation and maintenance work orders.

[0056] Furthermore, the reliability identification module 13 is used to execute the following method:

[0057] Retrieve the historical operation and maintenance database, and randomly extract the first set of operation and maintenance feature vectors from the multiple sets of operation and maintenance feature vectors. Among them, the historical operation and maintenance database includes multiple sets of historical operation and maintenance feature vectors of multiple historical operation and maintenance work orders and multiple reliable factors of historical operation and maintenance work orders; respectively, using any one of the operation and maintenance feature vectors in the first set of operation and maintenance feature vectors as an index, retrieve the multiple sets of historical operation and maintenance feature vectors of the multiple historical operation and maintenance work orders in the historical operation and maintenance database to obtain a first reliability identification sub-database. Among them, the first reliability identification sub-database includes multiple reliable factors of the first retrieved historical operation and maintenance work orders; conduct a tracking density analysis on the multiple reliable factors of the first retrieved historical operation and maintenance work orders, and determine the reliable factor of the first operation and maintenance work order according to the analysis result; conduct retrieval and tracking density analysis on the multiple sets of operation and maintenance feature vectors in the historical operation and maintenance database respectively to determine the reliable factors of the multiple operation and maintenance work orders.

[0058] Furthermore, the reliability identification module 13 is used to execute the following method:

[0059] Randomly extract a reliable factor of the first retrieved historical operation and maintenance work order from the multiple reliable factors of the first retrieved historical operation and maintenance work orders as the starting center; construct a starting neighborhood of the starting center based on a preset tracking bandwidth; use a weighted center iteration function to analyze and iterate the multiple reliable factors of the first retrieved historical operation and maintenance work orders and the starting center in the starting neighborhood to obtain an iterative weighted center, and then construct a neighborhood based on the iterative weighted center and conduct iteration again until the distance between the iterative weighted centers obtained in two adjacent iterations is less than or equal to a preset distance threshold when the preset iteration stop condition is met, and obtain the target weighted center. Among them, the preset iteration stop condition is that the distance between the iterative weighted centers obtained in two adjacent iterations is less than or equal to the preset distance threshold; use the reliable factor of the first retrieved historical operation and maintenance work order corresponding to the target weighted center as the reliable factor of the first operation and maintenance work order.

[0060] Furthermore, the reliability identification module 13 is used to execute the following method:

[0061] Construct the weighted center iteration function, where the weighted center iteration function is: where Point(x) is the iterative weighted center, N(x) is the multiple reliable factors of the first retrieved historical operation and maintenance work orders in the starting neighborhood, x i is the i-th reliable factor of the first retrieved historical operation and maintenance work orders among the multiple reliable factors of the first retrieved historical operation and maintenance work orders, x is the starting center, and K(x i -x) is a weight kernel function constructed based on the Gaussian function.

[0062] Furthermore, the optimization module 16 is used to execute the following method:

[0063] Discriminate the multiple initial device scheduling schemes using the primary scheduling constraint, eliminate the schemes that do not meet the primary scheduling constraint, and obtain multiple screened device scheduling schemes; traverse the multiple screened device scheduling schemes for scheduling fitness analysis to obtain multiple screened scheduling fitnesses; select the screened device scheduling corresponding to the maximum value among the multiple screened scheduling fitnesses as the optimization target, and adjust the remaining multiple screened device scheduling schemes according to a preset adjustment scheme to obtain multiple adjusted device scheduling schemes, where the preset adjustment scheme is to randomly adjust the scheduling order in the multiple screened device scheduling schemes with the device scheduling order in the optimization target as the goal; update the optimization target according to the multiple adjusted device scheduling schemes, and after multiple update iterations, use the adjusted device scheduling scheme corresponding to the optimization target obtained in the last iteration as the target intelligent scheduling scheme.

[0064] Further, the optimization module 16 is used to execute the following method:

[0065] Traverse the multiple adjusted device scheduling schemes for scheduling fitness analysis to obtain multiple adjusted scheduling fitnesses; determine whether there is any value greater than or equal to the screened scheduling fitness corresponding to the optimization target among the multiple adjusted scheduling fitnesses. If so, use the adjusted device scheduling scheme corresponding to the maximum value among the multiple adjusted scheduling fitnesses as the new optimization target and continue the update iteration; if not, add the adjustment method for obtaining the multiple adjusted device scheduling schemes to the tabu adjustment methods, and do not allow it to be used again within the preset tabu iteration times, and based on the optimization target, adjust the remaining multiple screened device scheduling schemes according to the preset adjustment scheme except for the tabu adjustment method, and update and iterate the optimization target according to the adjustment result.

[0066] Further, the operation and maintenance vector extraction module 12 is used to execute the following method:

[0067] Obtain multiple sample operation and maintenance work orders and multiple sample operation and maintenance feature vector sets as a sample data set; perform supervised training on the network layer constructed based on the feedforward neural network according to the sample data set to learn the mapping relationship between the operation and maintenance work orders and the operation and maintenance feature vectors until the training converges to obtain a trained vector extraction network layer; use the vector extraction network layer to identify the feature vectors of the multiple operation and maintenance work orders to obtain the multiple operation and maintenance feature vector sets.

[0068] It should be noted that the above sequence of embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above describes specific embodiments of this specification. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0069] The above are only the preferred embodiments of the present application, and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

[0070] This specification and the drawings are only exemplary descriptions of the present application, and are considered to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. An intelligent scheduling method for photovoltaic power station operation and maintenance equipment, characterized in that: The method comprises: Acquire multiple operation and maintenance work orders of the target photovoltaic power station, wherein each operation and maintenance work order includes a matching operation and maintenance equipment type and a matching operation and maintenance equipment working time; Traversing the multiple operation and maintenance work orders, extracting operation and maintenance vectors, and obtaining multiple operation and maintenance feature vector sets; Based on the multiple operation and maintenance feature vector sets, the multiple operation and maintenance work orders are identified for execution reliability, and multiple operation and maintenance work order reliability factors are determined; Acquire the operation and maintenance equipment resources of the target photovoltaic power station, and take the scheduling scheme satisfying the operation and maintenance equipment resources as the primary scheduling constraint; Based on the size of the reliability factors of the multiple operation and maintenance work orders, combined with the matching operation and maintenance equipment type and the matching operation and maintenance equipment working hours corresponding to each of the multiple operation and maintenance work orders, an operation and maintenance work order processing plan is identified to determine multiple initial equipment scheduling plans; Performing constraint optimization on the multiple initial equipment scheduling solutions according to the primary scheduling constraints to obtain a target intelligent scheduling solution; The operation and maintenance equipment resources of the target photovoltaic power station are called according to the target intelligent scheduling plan to complete the processing of the multiple operation and maintenance work orders.

2. The intelligent scheduling method for photovoltaic power station operation and maintenance equipment according to claim 1, characterized in that: include: Retrieving a historical operation and maintenance database, and randomly extracting a first operation and maintenance feature vector set from the multiple operation and maintenance feature vector sets, wherein the historical operation and maintenance database includes multiple historical operation and maintenance feature vector sets of multiple historical operation and maintenance work orders and multiple historical operation and maintenance work order reliability factors; Retrieve multiple historical operation and maintenance feature vector sets of multiple historical operation and maintenance work orders of the historical operation and maintenance database using any one of the first operation and maintenance feature vector sets as an index, and obtain a first reliability identification sub-database, wherein the first reliability identification sub-database includes multiple first retrieved historical operation and maintenance work order reliability factors; Conducting intensive tracking analysis on the reliability factors of multiple first searched historical operation and maintenance work orders, and determining the reliability factor of the first operation and maintenance work order according to the analysis results; The multiple operation and maintenance feature vector sets are searched and traced for intensive analysis in the historical operation and maintenance database to determine the multiple operation and maintenance work order reliability factors.

3. The intelligent scheduling method for photovoltaic power station operation and maintenance equipment according to claim 2, characterized in that: Conduct intensive tracking analysis on the reliability factors of multiple first search history operation and maintenance work orders, and determine the reliability factor of the first operation and maintenance work order based on the analysis results, including: Randomly extracting a first retrieval history maintenance work order reliability factor from the plurality of first retrieval history maintenance work order reliability factors as a starting center; Constructing a starting neighborhood of the starting center based on a preset tracking bandwidth; A weighted center iteration function is used to analyze and iterate the multiple first search history operation and maintenance work order reliability factors in the starting neighborhood and the starting center to obtain an iterative weighted center, and a neighborhood is constructed again based on the iterative weighted center and iterated until a preset iteration stop condition is met to obtain a target weighted center, wherein the preset iteration stop condition is that the distance between the iterative weighted centers obtained in two adjacent iterations is less than or equal to a preset distance threshold; The first searched historical operation and maintenance work order reliability factor corresponding to the target weighted center is used as the first operation and maintenance work order reliability factor.

4. The intelligent dispatching method for photovoltaic power station operation and maintenance equipment according to claim 3, characterized in that: include: Construct the weighted center iteration function, wherein the weighted center iteration function is: Point(x) is the iterative weighted center, N(x) is the reliability factor of multiple first searched historical maintenance work orders in the starting neighborhood, and x i is the i-th first retrieval history operation and maintenance work order reliability factor among multiple first retrieval history operation and maintenance work order reliability factors, x is the starting center, K(x i -x) is a weight kernel function built based on the Gaussian function.

5. The intelligent scheduling method for photovoltaic power station operation and maintenance equipment according to claim 1, characterized in that: include: Using the primary scheduling constraint to discriminate the multiple initial equipment scheduling solutions, eliminating solutions that do not meet the primary scheduling constraint, and obtaining multiple screening equipment scheduling solutions; Traversing the plurality of screening equipment scheduling plans to perform scheduling fitness analysis to obtain a plurality of screening scheduling fitnesses; The screening equipment scheduling corresponding to the maximum value among the multiple screening scheduling fitnesses is selected as the optimization target, and the remaining multiple screening equipment scheduling schemes are adjusted according to the preset adjustment scheme to obtain multiple adjusted equipment scheduling schemes, wherein the preset adjustment scheme is to take the equipment scheduling sequence in the optimization target as the target, and randomly adjust the scheduling sequence in the multiple screening equipment scheduling schemes; The optimization target is updated according to the multiple adjustment equipment scheduling schemes. After multiple update iterations, the adjustment equipment scheduling scheme corresponding to the optimization target obtained in the last iteration is used as the target intelligent scheduling scheme.

6. The intelligent dispatching method for photovoltaic power station operation and maintenance equipment according to claim 5, characterized in that: include: Traversing the plurality of adjustment equipment scheduling schemes to perform scheduling fitness analysis to obtain a plurality of adjustment scheduling fitnesses; Determine whether there is a screening scheduling fitness greater than or equal to the optimization target among the multiple adjustment scheduling fitnesses, and if so, take the adjustment device scheduling scheme corresponding to the maximum value among the multiple adjustment scheduling fitnesses as the new optimization target, and continue to update and iterate; If not, the adjustment method of the multiple adjustment equipment scheduling schemes will be added to the taboo adjustment method, and will not be allowed to be used again within the preset taboo iteration number. Based on the optimization objective, the remaining multiple screening equipment scheduling schemes will be adjusted again according to the preset adjustment schemes except the taboo adjustment method, and the optimization objective will be updated and iterated according to the adjustment results.

7. The intelligent scheduling method for photovoltaic power station operation and maintenance equipment according to claim 1, characterized in that: include: Acquire multiple sample operation and maintenance work orders and multiple sample operation and maintenance feature vector sets as sample data sets; Performing supervised training on a network layer constructed based on a feedforward neural network according to the sample data set, learning a mapping relationship between an operation and maintenance work order and an operation and maintenance feature vector, until the training reaches convergence, thereby obtaining a trained vector extraction network layer; The vector extraction network layer is used to perform feature vector recognition on the multiple operation and maintenance work orders to obtain the multiple operation and maintenance feature vector sets.

8. An intelligent dispatching platform for photovoltaic power station operation and maintenance equipment, characterized in that: The platform is used to implement the intelligent scheduling method for photovoltaic power station operation and maintenance equipment according to any one of claims 1 to 7, and comprises: Operation and maintenance work order acquisition module: obtains multiple operation and maintenance work orders of the target photovoltaic power station, where each operation and maintenance work order includes the matching operation and maintenance equipment type and the matching operation and maintenance equipment working hours; An operation and maintenance vector extraction module: traverses the multiple operation and maintenance work orders, extracts operation and maintenance vectors, and obtains multiple operation and maintenance feature vector sets; Reliability identification module: performing execution reliability identification on the multiple operation and maintenance work orders based on the multiple operation and maintenance feature vector sets, and determining multiple operation and maintenance work order reliability factors; Constraint setting module: obtaining the operation and maintenance equipment resources of the target photovoltaic power station, and taking the scheduling scheme satisfying the operation and maintenance equipment resources as the primary scheduling constraint; Solution identification module: based on the size of the reliability factors of the multiple operation and maintenance work orders, combined with the matching operation and maintenance equipment type and matching operation and maintenance equipment working hours corresponding to each of the multiple operation and maintenance work orders, identify the operation and maintenance work order processing solution and determine multiple initial equipment scheduling solutions; Optimization module: performing constraint optimization on the multiple initial equipment scheduling schemes according to the primary scheduling constraints to obtain a target intelligent scheduling scheme; Processing module: calling the operation and maintenance equipment resources of the target photovoltaic power station according to the target intelligent scheduling plan, and completing the processing of the multiple operation and maintenance work orders.

Images