Uninterruptible power operation intelligent control method, system and device, and storage medium

Abstract

The invention discloses an uninterruptible operation intelligent control method, system and device and a storage medium, and the method comprises the steps: obtaining an operation plan, and determining an operation task point; monitoring data of the operation task points are collected in real time, risk assessment is carried out, the risk condition of each operation task point in the next preset time period is predicted, a first emergency plan is correspondingly matched, and isolation measures are executed; in the next preset time period, risk assessment is conducted on each unisolated operation task point, the risk condition of the unisolated operation task point at the current moment is obtained, whether the isolation requirement is met or not is judged, statistics is conducted on the real-time unisolated operation task points according to the judgment result, and the optimal operation path is obtained in real time through the path planning model; and acquiring the operation condition of the operation task point in real time, updating the operation task point in the operation plan, and repeating the steps after the next preset time period until the operation plan is completed. The safety of the non-power-cut operation process can be guaranteed, and the overall operation efficiency is improved.

Description

Technical Field

[0002] This application relates to the technical field of power live working control, and specifically relates to an intelligent control method, system, device and storage medium for live working. Background Art

[0003] In the field of power system operation and maintenance, ensuring power supply continuity and security has always been one of the key tasks. With the continuous growth of social power demand, the traditional power outage operation mode can no longer meet the requirements of the efficient and reliable operation of modern power grids. To address this challenge, the industry has gradually promoted live working technology. Through advanced equipment and methods, maintenance, transformation and other work can be completed while the power grid remains in operation, significantly improving power supply reliability and work efficiency.

[0004] However, with the increasingly complex operation environment, the limitations of traditional technical means have gradually emerged and are difficult to meet higher-level requirements. Currently, the technical means for live working mainly include the operation of traditional tools dominated by manual experience, operation specifications based on fixed processes, and the support of some preliminary integrated automated equipment. For example: Operators usually rely on portable detection instruments for on-site monitoring and execute tasks in combination with pre-established operating procedures; at the same time, some areas have begun to introduce simple automated devices to assist in completing some repetitive work, such as using robotic arms for basic maintenance, etc. Although the above technical means can maintain power supply to a certain extent, they rely more on manual judgment and experience operation, lack the ability to finely control operation risks, and show obvious limitations especially when facing complex operation environments; in addition, existing technologies are difficult to dynamically adjust operation strategies from the perspective of the overall operation task, comprehensively consider operation safety and operation efficiency, optimize the overall operation process, and improve the overall operation safety and operation efficiency.

[0005] Therefore, it is an urgent problem to be solved to provide an intelligent control method and system for live working, ensure the safety of the live working process and improve the overall operation efficiency, and realize the refined management of the live working process. Summary of the Invention

[0006] In order to ensure the safety of the live working process and improve the overall operation efficiency, and realize the refined management of the live working process, this application provides an intelligent control method, system, device and storage medium for live working.

[0007] In a first aspect, this application provides an intelligent control method for live working, including: Obtain an operation plan and determine operation task point information; For each operation task point, monitor data of the operation task point in real time, and input them into a pre-constructed first risk assessment model respectively to predict and obtain the risk type, risk value and risk level of each operation task point in the next preset time period; using deep learning technology, for each operation task point, according to the predicted risk type, risk value and risk level in the next preset time period, obtain and execute the first emergency plan in the next preset time period, and the first emergency plan includes: whether to take isolation measures, the scope and duration of taking isolation measures; In the next preset time period, for each unisolated operation task point, monitor data of the operation task point in real time, and input them into a pre-constructed second risk assessment model respectively to obtain the risk type, risk value and risk level of the unisolated operation task point at the current moment and determine whether the isolation requirement is met, and count the unisolated operation task points in real time according to the judgment result; based on the unisolated operation task points in real time, use the first operation path planning model to obtain the optimal operation path in real time to guide the operation; the first operation path planning model selects a multi-objective optimization model, and the parameters of the corresponding objective function include the risk values of the unisolated operation task points in real time and the distance values between the unisolated operation task points in real time, and solve to obtain the optimal operation path with the minimum risk value and the shortest operation path distance; Obtain the operation status of the operation task point in real time and update the operation task points in the operation plan, and repeat the risk assessment of the operation task point, the isolation measures in the matching stage, obtaining the unisolated operation task points in real time and obtaining the optimal path after the next preset time period until the operation of all operation task points is completed.

[0008] By adopting the above scheme, monitor the data of the operation task point in real time and conduct risk assessment of each operation task point, obtain the emergency plan and pre-isolate the operation task points with high risks, combine with the operation task points that need to be isolated judged in real time, better ensure the operation safety, and further plan the optimal operation route with low risks and short paths on the basis of ensuring safety to guide the operation, improve the operation efficiency, optimize the operation process, and realize the refined management of the live working process.

[0009] Preferably, it further includes: For each operation task point, monitor data of the operation task point in real time, and input them into a pre-constructed third risk assessment model respectively to predict and obtain the risk type, risk value and risk level of each operation task point at each moment in the next preset time period; Using deep learning technology, for each operation task point, according to the predicted risk type, risk value and risk level at each moment in the next preset time period, obtain the second emergency plan in the next preset time period and replace the obtained first emergency plan for execution; the second emergency plan includes: whether to take isolation measures, the isolation method of taking phased isolation measures, the isolation scope and the isolation duration.

[0010] By adopting the above solution, monitoring data is collected in real time, and the risk type, risk value, and risk level at each moment in the next preset time period are predicted. The risk changes at each operation task point are determined, and a more accurate emergency plan is obtained. Compared with only conducting risk assessment and emergency treatment for the entire preset time period, it can respond to risk changes more timely, reduce unnecessary isolation operations, improve operation efficiency, and effectively reduce potential safety hazards.

[0011] Preferably, it further includes: When determining the operation task point information, the operation resources allocated by the operation plan are obtained synchronously; For each operation task point, the operation resources required for taking isolation measures in the emergency plan for the next preset time period obtained synchronously are obtained, and the first remaining operation resources after the execution of the emergency plan are calculated; During the next preset time period, for each unisolated operation task point, when the risk type, risk value, and risk level of the unisolated operation task point are obtained and it is determined that the isolation requirement is met, the operation resources required for the preset isolation measure matching the risk type, risk value, and risk level of the operation task point are determined, and the second remaining operation resources are calculated; for each unisolated operation task point, when the risk type, risk value, and risk level of the unisolated operation task point are obtained and it is determined that the isolation requirement is not met, the range of operation resources required for the operation task point is determined; Based on the real-time unisolated operation task points, the second operation path planning model is selected to replace the first operation path planning model to obtain the optimal operation path to guide the operation; the second operation path planning model selects a multi-objective optimization model, and the parameters of the corresponding objective function include the risk values of the real-time unisolated operation task points, the distance values between the real-time unisolated operation task points, and the range of operation resources required for the real-time unisolated operation task points. The constraint conditions include: the paths between operation task points are connected, the risk value of a single operation task point is less than the risk threshold, and the sum of the operation resources required for the real-time unisolated operation task points is less than the second remaining operation resources. The optimal operation path with the minimum risk, the shortest operation path, and the lowest operation resource consumption is obtained by solving.

[0012] By adopting the above solution, the second operation path planning model is introduced, which not only considers the risk level and operation time consumption of the operation, but also takes resource consumption as an important optimization goal, thereby effectively reducing resource waste and improving the overall operation efficiency.

[0013] Preferably, it further includes: According to the preset priority evaluation index, the priority of each work task point in the daily work plan in the work plan is quantitatively scored to obtain the priority classification of each work task point in the daily work plan, including: the first priority corresponding to the work that must be completed on the same day, the second priority corresponding to the work that is completed on the same day to the greatest extent, and the third priority that allows the work to be delayed until the next day; Based on real-time non-isolated work task points, an optimized first work path planning model is used to replace the first work path planning model to obtain the optimal work path to guide the work; the optimized first work path planning model is provided with constraints that the path generated on the day must include all the first-priority work task points, the path generated on the day includes the second-priority work task points to the greatest extent, and the path generated on the day includes the third-priority work task points under the condition that the time of the day permits, and when solving the multi-objective optimization model, a genetic algorithm based on chromosome encoding for the first-priority work task points and the second-priority work task points and a taboo search algorithm based on dynamic insertion of the third-priority work task points are selected to complete the acquisition of the optimal path.

[0014] By adopting the above scheme, priority evaluation indicators are introduced to quantitatively evaluate the priority of the work task points. The genetic algorithm and taboo search algorithm are combined to generate the optimal work path with the lowest risk, shortest path and lowest resource consumption under the priority constraint conditions, ensuring that the key tasks that must be completed on the day are given priority, taking into account the reasonable arrangement of the second and third priority work tasks, and effectively balancing work efficiency and flexibility.

[0015] Preferably, it also includes: For each work task point, after obtaining the first emergency plan for the next preset time period and before executing the first emergency plan, determine the other covered work task points according to the isolation measures scope of the current work task point, query the building information between the current work task point and each of the other covered work task points, and determine whether there is an obstacle building serving as an isolation barrier. If so, eliminate the other covered work task points that have an obstacle building serving as an isolation barrier between the corresponding work task point and the corresponding work task point.

[0016] By adopting the above scheme, before executing the first emergency plan, the building information of other work task points within the scope of the isolation measures is queried, and the work task points affected by the obstacle buildings are identified and eliminated, thereby avoiding the failure of isolation measures or waste of resources caused by obstacle buildings, and improving the safety and efficiency of non-stop operations.

[0017] Preferably, it also includes: After determining the information of the operation task points, first cluster the positions of each operation task point, divide the operation areas according to the clustering, and for the operation task points in each area, continue to complete the subsequent steps for each operation task point, predict and obtain the risk type, risk value and risk level of each operation task point in the next preset time period, obtain the first emergency plan for the next preset time period and execute it, and obtain the optimal path.

[0018] By adopting the above solution, considering the situation where the number of operation task points is large and scattered, clustering analysis is carried out on the positions of each operation task point, the operation areas are reasonably divided, effectively reducing the resource waste and time consumption caused by cross-regional operations, and improving the operation efficiency.

[0019] Preferably, the setting of the next preset time period includes: Obtain the number of historical frequencies of the risk levels of each operation task point in the operation plan being at the high risk level that is greater than the preset frequency; when the obtained number is greater than the preset number, set the duration of the next preset time period to the first preset duration, otherwise, set the duration of the next preset time period to the second preset duration, and the second preset duration is greater than the preset duration.

[0020] By adopting the above solution, the duration of the next preset time period is dynamically adjusted according to the historical risk data of the operation task points, making the risk assessment and emergency plan formulation more accurate, and improving the sensitivity and response ability to high-risk situations.

[0021] In a second aspect, the present application provides a live working intelligent control system, including: An operation data acquisition module, configured to acquire an operation plan and determine the information of the operation task points; An operation isolation control module, configured to, for each operation task point, collect the monitoring data of the operation task point in real time, and respectively input it into a pre-constructed first risk assessment model to predict and obtain the risk type, risk value and risk level of each operation task point in the next preset time period; using deep learning technology, for each operation task point, according to the predicted risk type, risk value and risk level of the next preset time period, correspondingly obtain the first emergency plan for the next preset time period and execute it, and the first emergency plan includes: whether to take isolation measures, the scope and duration of taking isolation measures; The operation path planning module is used to, within the next preset time period, for each unisolated operation task point, collect the monitoring data of the operation task point in real time, and input them into the pre-constructed second risk assessment model respectively, to obtain the risk type, risk value and risk level of the unisolated operation task point at the current moment, and judge whether the isolation requirement is met, and count the unisolated operation task points in real time according to the judgment result; based on the unisolated operation task points in real time, use the first operation path planning model to obtain the optimal operation path in real time to guide the operation; the first operation path planning model selects a multi-objective optimization model, and the parameters corresponding to the objective function include the risk values of the unisolated operation task points in real time and the distance values between the unisolated operation task points in real time, and solve to obtain the optimal operation path with the smallest risk value and the shortest operation path distance; The operation content update module is used to obtain the operation situation of the operation task point in real time and update the operation task point in the operation plan, and repeat the risk assessment of the operation task point, matching the isolation measures in the stage, obtaining the unisolated operation task points in real time and obtaining the optimal path after the next preset time period until the operation of all operation task points is completed.

[0022] By adopting the above scheme, intelligent power-off-free operation control is realized, the safety of the power-off-free operation process is guaranteed, the overall operation efficiency is improved, and the refined management of the power-off-free operation process is realized.

[0023] In a third aspect, the present application provides a computer-readable storage medium, which includes a stored computer program, wherein, when the computer program runs, it controls the device where the computer-readable storage medium is located to execute the method as described above.

[0024] In a fourth aspect, the present application provides a computer device, which includes a memory, a processor, and a program stored on the memory and executable, and when the program is executed by the processor, it realizes the steps of the method as described above.

[0025] In summary, the present application has the following beneficial effects: 1. By collecting monitoring data in real time and combining deep learning technology for risk assessment, accurate prediction and control of the risks of operation task points are realized, so as to obtain corresponding emergency plans and pre-execute intelligent isolation measures, and combine real-time dynamics to judge whether to isolate, jointly ensuring the real-time operation safety under complex environmental conditions; at the same time, based on the operation path planning of the multi-objective optimization model, the operation path is dynamically adjusted, and the operation task is completed based on the optimal operation path, improving the operation efficiency; 2. Conduct a more detailed risk assessment of the operation tasks, and match more detailed emergency plans to improve the operation safety; consider the operation resources, and obtain the best operation path by comprehensively considering operation safety, operation distance and operation resources to guide the operation, realizing more refined operation process management; 3. Consider the priority of the job tasks, optimize the path model based on the constraints set by the priority of the job task points, combine the genetic algorithm and the tabu search algorithm to solve the optimal path, ensure the priority processing of key tasks, and improve the overall efficiency and safety of the job. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a flowchart of the intelligent control method for live working described in the specific embodiment; Figure 2 is a schematic structural diagram of the intelligent control system for live working described in the specific embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0028] As Figure 1 shown, the embodiments of the present application disclose an intelligent control method for live working, which specifically includes: S1. Obtain the job plan and determine the information of the job task points.

[0029] Specifically, for the live working plan in the power system, the plan content includes the information of the job task points (including but not limited to the location of the job task points, the equipment and surrounding conditions corresponding to the job task points, the correlation between the job task points and other job task points, the job content of the job task points, etc.), and allocate job resources (including but not limited to job tools, job personnel, job auxiliary tools, etc.).

[0030] In addition, considering that in the live working plan, the positions of the job task points are relatively concentrated and relatively dispersed. Since before the next operation, the positions of each job task point are preferentially clustered, and based on the clustering results, regional division is carried out, such as divided into area A, area B, and area C, and each job task point in each area is distinguished and numbered. For example, there are job task points numbered 001-006 in area A.

[0031] S2. For each job task point, collect the monitoring data of the job task point in real time and conduct risk assessment.

[0032] Specifically, for each job task point, the monitoring data of each job task point is collected through the deployed IoT sensors and SCADA system. Among them, the monitoring data includes: environmental data (temperature, humidity, etc.), operating parameter data (equipment status, voltage, current), personnel operation specification data, and operation and maintenance data (operation and maintenance personnel, operation and maintenance time, etc.).

[0033] Using deep learning technology, a risk assessment model for job task points is pre-constructed, including a first risk assessment model and a second risk assessment model. Among them, the input of the first risk assessment model is the monitoring data of the job task point, and the output is the risk type, risk value, and risk level of the job task point in the next preset period. It is trained and generated using the monitoring data of the job task point at any historical moment (any moment and the previous period), the risk type, risk value, and risk level of the job task point in the next preset period at any historical moment as training data. For each job task point, the monitoring data of the job task point is collected in real time and input into the pre-constructed first risk assessment model to predict and obtain the risk type, risk value, and risk level of the corresponding job task point in the next preset period.

[0034] Among them, the next preset period can be set manually according to experience. In order to make more detailed predictions for each job task point with frequent high risks, the historical frequency of the risk level of each job task point in the job plan being at a high risk level greater than the preset frequency is selected and obtained. When the obtained quantity is greater than the preset quantity, it indicates that there are relatively many job task points in the current plan that are prone to high risk levels. Therefore, more detailed predictions are made and isolation measures are implemented, and the duration of the next preset period is set to the first preset duration. Otherwise, the duration of the next preset period is set to the second preset duration, and the second preset duration is greater than the preset duration.

[0035] In addition, if there is a situation where the job task points in the job plan are divided into regions, multiple first risk assessment models can be set. For the monitoring data of each job task point in each specific region, it is input into the corresponding first risk assessment model of the region, and the prediction of the risk type, risk value, and risk level of each job task point in the next preset period is continued for each job task point within the specific region.

[0036] S3. Using deep learning technology, for each job task point, according to the predicted risk type, risk value, and risk level in the next preset period, the first emergency plan for the next preset period is obtained and executed accordingly.

[0037] Specifically, considering that some job task points may have high risks, in order to ensure job safety, it is necessary to pre-match emergency plans for these job task points, such as taking isolation measures, and corresponding job resources are allocated for isolation processing and operation recovery processing of the job task points. The specific isolation scope and isolation duration of these job task points need to be set according to the specific risk type and level.

[0038] Therefore, deep learning technology can be utilized to construct a first emergency plan acquisition model. The input of the first emergency plan acquisition model is the risk type, risk value, and risk level predicted for the next preset time period corresponding to each operation task, and the output is the first emergency plan for the next preset time period. The first emergency plan includes: whether to take isolation measures, the ways of taking isolation measures (such as physical isolation, logical isolation, dynamic flow limiting, etc.), scope and duration, and the operation resources required for taking isolation measures, etc. It is trained and generated by using the risk type, risk value, and risk level within the preset time period of historical operation task points and the content of the first emergency plan matched based on expert experience for the preset time period of historical operation task points as training data.

[0039] Execute the first emergency plan for the next preset time period obtained for a single operation task point until an isolation operation instruction is generated to indicate the allocation of required operation resources, complete the isolation processing of the operation task point, or generate an instruction indicating no need for isolation to indicate no operation is required.

[0040] In addition, correspondingly, if there is a situation of dividing operation task points in the operation plan into regions, multiple first emergency plan acquisition models can be set. For the risk type, risk value, and risk level of each operation task point in the next preset time period in each specific region, input them into the corresponding first emergency plan acquisition model of the region, and continue to complete the acquisition of the first emergency plan for each operation task point within the specific region.

[0041] S4. In the next preset time period, obtain the real-time unisolated operation task points, and use the first operation path planning model to obtain the optimal operation path in real time to guide the operation.

[0042] Specifically, in the next preset time period, pre-obtain the real-time unisolated operation task points.

[0043] First, after the execution of the emergency plan mentioned above, some operation task points have been isolated, including the operation task points determined to need isolation processing, and also including the operation task points within the isolation range of the operation task points determined to need isolation processing. For example, for task point 4, it is predicted that isolation is required in the next preset time period, and the isolation range is 50m. Task point 6 is also within this isolation range. Then, after obtaining and executing the emergency plan for each operation task point, count the real-time unisolated operation task points.

[0044] Secondly, considering each job task point where isolation has not been performed in advance prediction, in order to further ensure the safety of the job task point, in the next preset time period, for each unisolated job task point, the monitoring data of the job task point is collected in real time and input into the pre-constructed second risk assessment model respectively to obtain the risk type, risk value and risk level of the unisolated job task point at the current moment; the second risk assessment model adopts a neural network model, the input of the model is the monitoring data of the job task point, and the output is the risk type and risk level of the job task point in the current time period, which is trained and generated by using the monitoring data of the job task point at any historical moment, the risk type, risk value and risk level of the job task point marked by experts at any historical moment as training data.

[0045] Based on obtaining the risk type, risk value and risk level of the unisolated job task point at the current moment for each unisolated job task point in the next preset time period, it is correspondingly judged whether the isolation requirement is met. For job task points of different risk types, when the risk level reaches the corresponding matching risk level, it is judged that isolation is required. For example, when there is partial discharge at the job task point and it reaches the medium risk level, isolation is required; according to the judgment result, the unisolated job task points in real time are counted, that is, an isolation operation instruction is generated for the job task points judged to require isolation and the execution of isolation is instructed, and the remaining unisolated job task points at the current moment are counted.

[0046] In the next preset time period, based on the unisolated job task points in real time, the optimal job path is obtained in real time by using the first job path planning model to guide the operation, that is, the operation personnel are guided in real time to complete the operation of the job task points according to the job path, which can not only ensure the safety of the operation personnel but also save the path that the operation personnel need to travel as much as possible and improve the operation efficiency. Among them, the first job path planning model selects a multi-objective optimization model, and the parameters of the corresponding objective function include the risk values of the unisolated job task points in real time and the distance values between the unisolated job task points in real time, and the optimal job path with the smallest risk value and the shortest job path distance is obtained by solving; among them, the optimization objectives of the objective function include: the minimum comprehensive risk value of the path and the shortest total job path distance, and the constraints include: the path between job task points is connected, and the risk value of a single job task point is less than the risk threshold; it is optional to use the improved A * algorithm or genetic algorithm to solve the objective function to obtain the optimal job path.

[0047] In addition, correspondingly, if there is a situation of dividing the job task points in the job plan into regions, it is necessary to obtain the optimal path for the unisolated job task points in the specific region in real time.

[0048] S5. Update the job task points in the job plan, and repeat the risk assessment of job task points, matching-stage isolation measures, obtaining real-time unisolated job task points, and obtaining the optimal path after the next preset time period until the operations of all job task points are completed.

[0049] Specifically, within the next preset time period, generate an optimal job path in real time to guide the operator to complete the operations of job task points. Correspondingly, obtain the operation status of job task points in real time through the SCADA system, mark the job task points that have completed the operations, eliminate the job task points that have completed the operations, and update the job task points in the job plan. Repeat the risk assessment of job task points, matching-stage isolation measures, obtaining real-time unisolated job task points, and obtaining the optimal path after the next preset time period until the operations of all job task points are completed.

[0050] In a specific embodiment, in order to achieve refined prediction and dynamic response to the risks of job task points, the method further includes: For each job task point, monitor data of the job task point is collected in real time and input into a pre-constructed third risk assessment model respectively to predict and obtain the risk type, risk value, and risk level of each job task point at each moment in the next preset time period. Among them, the input of the third risk assessment model is the monitor data of the job task point, and the output is the risk type, risk value, and risk level of each job task point at each moment in the next preset time period, which is trained and generated by using the monitor data of the job task point at any historical moment (any moment and the previous period) and the risk type, risk value, and risk level of each job task point at each moment in the next preset time period at any historical moment as training data.

[0051] For each job task point, monitor data of the job task point is collected in real time and input into a pre-constructed third risk assessment model to predict and obtain the risk type, risk value, and risk level of the corresponding job task point at each moment in the next preset time period.

[0052] Considering that the risk type, risk value, and risk level of each moment in the predicted next preset time period can be obtained, the risk change of each job task point in the next preset time period can be obtained. For example, the risk value of partial discharge gradually decreases, and correspondingly, unified isolation is not required. Along with the decrease of the risk value, the isolation range is dynamically adjusted. Therefore, the deep learning technology can be used to construct a second emergency plan acquisition model. The input of the second emergency plan acquisition model is the risk type, risk value, and risk level of each moment in the predicted next preset time period corresponding to each job task, and the output is the second emergency plan for the next preset time period, which replaces the obtained first emergency plan and is executed. The second emergency plan includes: whether to take isolation measures, the isolation method of taking phased isolation measures, the isolation scope and duration, the operation resources required for taking phased isolation measures, etc.; among them, the phased isolation measures are accurate to each phased moment of each next preset time period; the second emergency plan content matched based on expert experience for the risk type, risk value and risk level of each moment in a period of time of historical operation task points and the risk type, risk value and risk level of historical operation task points at each moment in a period of time is used as training data for training and generation.

[0053] In a specific embodiment, in addition to the size of the operation risk and the distance of the operation path, the operation resources are often also the most important influencing factors for completing the operation task. In order to improve the operation resource utilization rate while improving the operation efficiency, so as to realize the refined management and optimized allocation of the operation process, the method further includes: When determining the operation task point information, synchronously obtain the operation resources allocated by the operation plan.

[0054] For each operation task point, synchronously obtain the operation resources required for taking isolation measures in the emergency plan of the next preset time period, and calculate the first remaining operation resources after the execution of the emergency plan = the operation resources allocated by the operation plan - the sum of the operation resources required for taking isolation measures in the emergency plan of the next preset time period corresponding to the operation task point.

[0055] Within the next preset time period, for each unisolated operation task point, obtain the risk type, risk value and risk level of the unisolated operation task point and when it is determined that the isolation requirement is met, determine the operation resources required for the preset isolation measure matched with the risk type, risk value and risk level of the operation task point, and calculate and obtain the second remaining operation resources = the first remaining resources - the sum of the operation resources required for the preset isolation measure matched with the risk type, risk value and risk level of each operation task point; for each unisolated operation task point, obtain the risk type, risk value and risk level of the unisolated operation task point and when it is determined that the isolation requirement is not met, determine the range of operation resources required for the operation task point. Based on the real-time unisolated operation task points, select to use the second operation path planning model to replace the first operation path planning model to obtain the optimal operation path to guide the operation.

[0056] Specifically, the second operation path planning model selects a multi-objective optimization model. The parameters corresponding to the objective function include the risk values of the real-time unisolated operation task points, the distance values between the real-time unisolated operation task points, and the range of operation resources required by the real-time unisolated operation task points. The constraint conditions include: the paths between operation task points are connected, the risk value of a single operation task point is less than the risk threshold, and the sum of the operation resources required by the real-time unisolated operation task points is less than the second remaining operation resources; the optimal operation path with the minimum risk, the shortest operation path, and the lowest operation resource consumption is obtained by solving.

[0057] In a specific embodiment, path planning optimization is performed considering the priority of operation tasks to ensure that operation task points with a higher priority level are completed first, effectively balancing operation efficiency and flexibility; the method further includes: According to the preset priority evaluation index, a quantitative score is given to the priority of each operation task point in the daily operation plan. The priority evaluation index includes the magnitude of the risk value of the operation task point, the importance of the operation task point in the entire power topology network, etc.

[0058] The priority classification of each operation task point in the daily operation plan is obtained, including: the first priority P1 corresponding to the operation that must be completed on the same day, the second priority P2 corresponding to the operation that is completed on the same day to the greatest extent, and the third priority P3 corresponding to the operation that is allowed to be postponed until the next day.

[0059] Based on the real-time unisolated operation task points, the optimized first operation path planning model is used to replace the first operation path planning model to obtain the optimal operation path to guide the operation. Among them, the parameters corresponding to the objective function of the optimized first operation path planning model include the risk values of the real-time unisolated operation task points, the distance values between the real-time unisolated operation task points, the priorities of the real-time operation task points, etc. In addition to the constraint condition that the paths between operation task points are connected and the risk value of a single operation task point is less than the risk threshold, the constraint conditions also include that the path generated on the same day must include all operation task points of the first priority, the path generated on the same day includes the operation task points of the second priority to the greatest extent, and the path generated on the same day includes the operation task points of the third priority under the condition that the time on the same day permits; the optimization objectives of the corresponding function include: the minimum comprehensive risk value of the path and the shortest total operation path distance.

[0060] When solving the objective function corresponding to the multi-objective optimization model, a genetic algorithm based on chromosome coding for the operation task points of the first priority and the operation task points of the second priority and a tabu search algorithm based on dynamically inserting the operation task points of the third priority are selected to obtain the optimal path. The specific steps include: Using a genetic algorithm as the main path planning, chromosome encoding is performed according to the priorities of unisolated job task points. Each chromosome represents a possible job path plan, which consists of a series of job task points. When encoding, consider assigning higher weights or more forward positions to task points with higher priorities, so that job task points at the P1 level are prioritized, such as: P1-3, P1-5, P2-2, P2-7, Depot (starting / ending point); generating an initial population, including: clustering the positions of all P1 job task points to generate an initial path, randomly inserting P2 job task points, and ensuring that there are no conflicts in time windows and resources to generate the population size; calculating the fitness function. For each chromosome (i.e., each job path plan), calculate its risk value and path length, and evaluate its fitness based on these two metrics; among them, the fitness can also set penalty terms, such as: time window violation penalty term: the number of overtime tasks multiplied by the penalty coefficient + resource conflict penalty term: the same device is repeatedly assigned multiplied by the number of conflicts, etc.; selection operation: according to the fitness evaluation results, select a part of the excellent chromosomes to enter the next generation, such as using roulette wheel selection method, tournament selection method, etc.; crossover operation: perform a crossover operation on the selected chromosomes to generate new job path plans, and single-point crossover, multi-point crossover or uniform crossover can be selected; mutation operation: perform a mutation operation on the newly generated chromosomes, and bit inversion, bit insertion or bit deletion can be selected; repeat the fitness evaluation, selection, crossover and mutation operations until the termination condition is met (such as reaching the maximum number of iterations or finding the optimal solution that meets the requirements); Using tabu search to dynamically insert P3 job task points as elastic path optimization, taking the path output at the end of the genetic algorithm iteration as the input, and initializing the tabu list and candidate list; the tabu list is used to record the inserted positions that have been tried to avoid repeated searches; evaluating the insertion effect: for each candidate insertion position, evaluate the changes in the risk and length of the path after inserting the third-priority task point; selecting the best insertion position: select the position that increases the path risk and length the least from the candidate list as the best insertion position; updating the path and tabu list: insert the third-priority task point at the best insertion position, and update the tabu list and candidate list; iterative search: repeat the above steps until the termination condition is met (such as reaching the maximum number of iterations or finding the optimal solution that meets the requirements); After the genetic algorithm and tabu search are completed, select the path with the highest fitness and meeting all constraint conditions as the optimal job path, decode the optimal job path plan to obtain the specific order and path of job task points, so as to guide the operators to perform actual job operations.

[0061] In a specific embodiment, in order to effectively avoid the problems of isolation measure failure or resource waste caused by obstacle buildings and improve the safety and efficiency of live working, the method further includes: For each job task point, after obtaining the first emergency plan for the next preset time period and before implementing the first emergency plan, determine other job task points covered according to the isolation measure scope of the current job task point; for example, when isolating job task point 4, determine that other job task points covered include job task 6. Query the building information or other substances between the current job task point and each job task point among the other covered job task points to determine whether there are obstacle buildings or other substances serving as isolation barriers; if so, remove the other covered job task points with obstacle buildings serving as isolation barriers between them and the corresponding job task points; that is, when there are obstacle buildings serving as isolation barriers between the current job task point and the other covered job task points, there is no need to specifically isolate and cover the current job task point 4 according to the isolation measure scope to job task point 6, and job task point 6 is relatively safe.

[0062] As Figure 2 shown, an intelligent control system for live working disclosed in an embodiment of the present application specifically includes: A job data acquisition module 101, configured to obtain a job plan and determine job task point information. A job isolation control module 102, configured to, for each job task point, collect monitoring data of the job task point in real time and input them into a pre-constructed first risk assessment model respectively to predict and obtain the risk type, risk value, and risk level of each job task point in the next preset time period; using deep learning technology, for each job task point, according to the predicted risk type, risk value, and risk level in the next preset time period, correspondingly obtain and execute the first emergency plan for the next preset time period, where the first emergency plan includes: whether to take isolation measures, the scope and duration of taking isolation measures. A job path planning module 103, configured to, within the next preset time period, for each unisolated job task point, collect monitoring data of the job task point in real time and input them into a pre-constructed second risk assessment model respectively to obtain the risk type, risk value, and risk level of the currently unisolated job task points at the current moment and determine whether the isolation requirement is met, and count the currently unisolated job task points according to the judgment result; based on the currently unisolated job task points, use a first job path planning model to obtain an optimal job path in real time to guide the job; the first job path planning model selects a multi-objective optimization model, and the parameters of the corresponding objective function include the risk values of the currently unisolated job task points and the distance values between the currently unisolated job task points, and solve to obtain the optimal job path with the minimum risk value and the shortest job path distance. The job content update module 104 is used to obtain the job status of job task points in real time and update the job task points in the job plan, and repeat the risk assessment of job task points, the matching stage isolation measures, obtaining the real-time unisolated job task points, and obtaining the optimal path after the next preset time period until the jobs of all job task points are completed.

[0063] In a specific embodiment, the job isolation control module 102 in the system is further configured to, for each job task point, collect the monitoring data of the job task point in real time and input them into a pre-constructed third risk assessment model respectively to predict and obtain the risk type, risk value, and risk level of each job task point at each moment in the next preset time period; using deep learning technology, for each job task point, according to the predicted risk type, risk value, and risk level of each moment in the next preset time period, obtain the second emergency plan in the next preset time period correspondingly and replace the obtained first emergency plan for execution; the second emergency plan includes: whether to take isolation measures, the isolation method of taking phased isolation measures, the isolation range, and the isolation duration.

[0064] In a specific embodiment, the job path planning module 103 in the system is further configured to, when determining the job task point information, synchronously obtain the job resources allocated by the job plan; for each job task point, synchronously obtain the job resources required for taking isolation measures in the emergency plan in the next preset time period, and calculate the first remaining job resources after the execution of the emergency plan; within the next preset time period, for each unisolated job task point, when obtaining the risk type, risk value, and risk level of the unisolated job task point and determining that the isolation requirement is met, determine the job resources required for the preset isolation measures matching the risk type, risk value, and risk level of the job task point, and calculate and obtain the second remaining job resources; for each unisolated job task point, when obtaining the risk type, risk value, and risk level of the unisolated job task point and determining that the isolation requirement is not met, determine the range of job resources required for the job task point; based on the real-time unisolated job task points, select to use the second job path planning model to replace the first job path planning model to obtain the optimal job path to guide the job; the second job path planning model selects a multi-objective optimization model, and the parameters corresponding to the objective function include the risk values of the real-time unisolated job task points, the distance values between the real-time unisolated job task points, and the range of job resources required for the real-time unisolated job task points, and the constraint conditions include: the paths between job task points are connected, the risk value of a single job task point is less than the risk threshold, and the sum of the job resources required for the real-time unisolated job task points is less than the second remaining job resources, and solve to obtain the optimal job path with the minimum risk, the shortest job path, and the lowest job resource consumption.

[0065] A specific embodiment, the operation path planning module 103 in the system is further configured to quantitatively score the priority of each operation task point in the daily operation plan according to a preset priority evaluation index, and obtain the priority classification of each operation task point in the daily operation plan, including: the first priority corresponding to the operation that must be completed on the current day, the second priority corresponding to the operation that is completed on the current day to the greatest extent, and the third priority allowing delay until the next day to complete the operation; based on the unisolated operation task points in real time, use the optimized first operation path planning model to replace the first operation path planning model to obtain the optimal operation path to guide the operation; the optimized first operation path planning model is provided with constraint conditions that the path generated on the current day must include all operation task points of the first priority, the path generated on the current day includes operation task points of the second priority to the greatest extent, and the path generated on the current day includes operation task points of the third priority under the condition that the time on the current day permits, and when solving the multi-objective optimization model, select a genetic algorithm based on chromosome coding for operation task points of the first priority and operation task points of the second priority and a tabu search algorithm based on dynamically inserting operation task points of the third priority to complete the acquisition of the optimal path.

[0066] A specific embodiment, the operation path planning module 103 is further configured to, for each operation task point, obtain the first emergency plan for the next preset time period, and before executing the first emergency plan, determine the other operation task points covered according to the isolation measure range of the current operation task point, query the building information between the current operation task point and each of the other operation task points covered, and determine whether there is an obstacle building serving as an isolation barrier. If so, remove the other operation task points covered that have an obstacle building serving as an isolation barrier between them and the corresponding operation task point.

[0067] A specific embodiment, the operation data acquisition module 101 is further configured to, after determining the operation task point information, first cluster the positions of each operation task point, divide the operation areas according to the clustering, and for the operation task points in each area, continue to complete the subsequent steps for each operation task point, predict and obtain the risk type, risk value and risk level of each operation task point in the next preset time period, obtain and execute the first emergency plan for the next preset time period, and obtain the optimal path.

[0068] The embodiment of the present application also discloses a computer-readable storage medium.

[0069] Specifically, this computer-readable storage medium stores a computer program that can be loaded and executed by a processor, such as the power-off-free operation intelligent control method as described above. This computer-readable storage medium includes, for example: various media that can store program codes such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks or optical discs.

[0070] The embodiments of the present application also disclose a computer device.

[0071] Specifically, the computer device includes a memory and a processor, and a computer program capable of being loaded and executed by the processor for the above-mentioned intelligent control method for uninterrupted operation is stored on the memory.

[0072] The above are all the preferred embodiments of the present application. The protection scope of the present application is not limited thereby. Any feature disclosed in this specification (including the abstract and drawings), unless specifically described, can be replaced by other equivalent or similar-purpose alternative features. That is, unless specifically described, each feature is only an example of a series of equivalent or similar features.

Claims

1. An intelligent control method for live working, characterized in that, Including: Obtain a job plan and determine job task point information; For each job task point, collect the monitoring data of the job task point in real time and input it into a pre-constructed first risk assessment model respectively to predict and obtain the risk type, risk value and risk level of each job task point in the next preset time period; Using deep learning technology, for each job task point, according to the predicted risk type, risk value and risk level in the next preset time period, obtain and execute the first emergency plan in the next preset time period correspondingly. The first emergency plan includes: whether to take isolation measures, the scope and duration of taking isolation measures; During the next preset time period, for each unisolated job task point, collect the monitoring data of the job task point in real time and input it into a pre-constructed second risk assessment model respectively to obtain the risk type, risk value and risk level of the unisolated job task point at the current moment and judge whether the isolation requirement is met. According to the judgment result, count the unisolated job task points in real time; Based on the unisolated job task points in real time, use the first job path planning model to obtain the optimal job path in real time to guide the operation; The first job path planning model selects a multi-objective optimization model, and the parameters corresponding to the objective function include the risk values of the unisolated job task points in real time and the distance values between the unisolated job task points in real time. The constraint conditions include: the paths between job task points are connected, and the risk value of a single job task point is less than the risk threshold. Solve to obtain the optimal job path with the smallest risk value and the shortest job path distance; Obtain the operation situation of the job task points in real time and update the job task points in the job plan. After the next preset time period, repeat the risk assessment of job task points, matching the isolation measures in the stage, obtaining the unisolated job task points in real time, and obtaining the optimal path until the operation of all job task points is completed.

2. The intelligent control method for live working according to claim 1, characterized in that Also including: For each job task point, collect the monitoring data of the job task point in real time and input it into a pre-constructed third risk assessment model respectively to predict and obtain the risk type, risk value and risk level of each job task point at each moment in the next preset time period; Using deep learning technology, for each job task point, according to the predicted risk type, risk value and risk level at each moment in the next preset time period, obtain the second emergency plan in the next preset time period correspondingly and replace the obtained first emergency plan for execution; The second emergency plan includes: whether to take isolation measures, the isolation method of taking phased isolation measures, the isolation scope and the isolation duration.

3. The intelligent control method for live working according to claim 1, wherein Also including: When determining the job task point information, synchronously obtain the job resources allocated by the job plan; For each job task point, synchronously obtain the job resources required for taking isolation measures in the emergency plan in the next preset time period, and calculate the first remaining job resources after the execution of the emergency plan; During the next preset time period, for each unisolated job task point, when obtaining the risk type, risk value and risk level of the unisolated job task point and judging that the isolation requirement is met, determine the job resources required for the preset isolation measures matching the risk type, risk value and risk level of the job task point, and calculate and obtain the second remaining job resources; For each unisolated job task point, when obtaining the risk type, risk value, and risk level of the unisolated job task point and determining that the isolation requirement is not met, determine the range of job resources required for the job task point; Based on the real-time unisolated job task points, select and use the second job path planning model to replace the first job path planning model to obtain the optimal job path to guide the job; The second job path planning model selects a multi-objective optimization model. The parameters corresponding to the objective function include the risk values of the real-time unisolated job task points, the distance values between the real-time unisolated job task points, and the range of job resources required for the real-time unisolated job task points. The constraint conditions include: the paths between job task points are connected, the risk value of a single job task point is less than the risk threshold, and the sum of the job resources required for the real-time unisolated job task points is less than the second remaining job resources. Solve to obtain the optimal job path with the minimum risk, the shortest job path, and the lowest job resource consumption.

4. The intelligent control method for live working according to claim 1, wherein It also includes: According to the preset priority evaluation index, quantitatively score the priority of each job task point in the daily job plan in the job plan to obtain the priority classification of each job task point in the daily job plan, including: the first priority corresponding to the job that must be completed on the same day, the second priority corresponding to the job that is completed on the same day to the greatest extent, and the third priority allowing delay until the next day to complete the job; Based on the real-time unisolated job task points, use the optimized first job path planning model to replace the first job path planning model to obtain the optimal job path to guide the job; the optimized first job path planning model has constraint conditions that the path generated on the same day must include all job task points with the first priority, the path generated on the same day includes job task points with the second priority to the greatest extent, and the path generated on the same day includes job task points with the third priority under the condition that the time on the same day permits. When solving the multi-objective optimization model, select a genetic algorithm based on chromosome coding for job task points with the first priority and job task points with the second priority and a tabu search algorithm based on dynamically inserting job task points with the third priority to complete the acquisition of the optimal path.

5. The intelligent control method for live working according to claim 1, characterized in that, It also includes: For each job task point, after obtaining the first emergency plan for the next preset time period, and before executing the first emergency plan, determine the other job task points covered according to the isolation measure range of the current job task point, query the building information between the current job task point and each of the other job task points covered, and determine whether there is an obstacle building serving as an isolation barrier. If so, eliminate the other job task points covered that have an obstacle building serving as an isolation barrier between them and the corresponding job task point.

6. The intelligent control method for live working according to claim 1, wherein It also includes: After determining the job task point information, first cluster the positions of each job task point, divide the job area according to the clustering, and for the job task points in each area, continue to complete the subsequent steps for each job task point, predict and obtain the risk type, risk value, and risk level of each job task point in the next preset time period, obtain and execute the first emergency plan for the next preset time period, and obtain the optimal path.

7. The intelligent control method for live working according to claim 1, wherein The setting of the next preset time period includes: Obtaining the number of historical frequencies of the risk levels of each job task point in the job plan being at a high risk level that is greater than a preset frequency; when the obtained number is greater than a preset number, the duration of the next preset time period is set to a first preset duration, otherwise, the duration of the next preset time period is set to a second preset duration, and the second preset duration is greater than the preset duration.

8. An intelligent control system for live working, characterized in that, Including: A job data acquisition module, configured to acquire a job plan and determine job task point information; A job isolation control module, configured to, for each job task point, collect monitoring data of the job task point in real time and input it into a pre-constructed first risk assessment model respectively to predict and obtain the risk type, risk value, and risk level of each job task point in the next preset time period; Using deep learning technology, for each job task point, according to the predicted risk type, risk value, and risk level in the next preset time period, obtain and execute the first emergency plan corresponding to the next preset time period, where the first emergency plan includes: whether to take isolation measures, the scope and duration of taking isolation measures; A job path planning module, configured to, within the next preset time period, for each unisolated job task point, collect monitoring data of the job task point in real time and input it into a pre-constructed second risk assessment model respectively to obtain the risk type, risk value, and risk level of the unisolated job task point at the current moment and determine whether the isolation requirement is met, and count the unisolated job task points in real time according to the judgment result; based on the unisolated job task points in real time, use the first job path planning model to obtain the optimal job path in real time to guide the job; the first job path planning model selects a multi-objective optimization model, and the parameters of the corresponding objective function include the risk values of the unisolated job task points in real time and the distance values between the unisolated job task points in real time, and the constraint conditions include: the paths between job task points are connected, and the risk value of a single job task point is less than the risk threshold, and solve to obtain the optimal job path with the smallest risk value and the shortest job path distance; A job content update module, configured to obtain the job situation of job task points in real time and update the job task points in the job plan, and repeat the risk assessment of job task points, matching stage isolation measures, obtaining unisolated job task points in real time, and obtaining the optimal path after the next preset time period until the jobs of all job task points are completed.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein when the computer program runs, it controls the device where the computer-readable storage medium is located to execute the method described in any one of claims 1 to 7.

10. A computer device, characterized in that, The computer device includes a memory, a processor, and a program stored and executable on the memory. When the program is executed by the processor, it implements the steps of the method described in any one of claims 1 to 7.

Images