Power grid dispatching virtual teacher training plan generation method and system
Through the virtual teacher designer and controller, a long process operation section and initial expected fault set are generated by the virtual teacher designer and controller, and probability fault calculation and evaluation are carried out, which solves the defects of the existing grid scheduling training system in scene simulation and interaction methods, and achieves high efficiency and individuality The integrated scheduling training improves the intelligence and practicality of the training.
Patent Information
- Application Number
- CN202411438652.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-15
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-10-15
AI Technical Summary
The existing power grid scheduling training system has defects in scenario simulation and interaction methods, and cannot effectively simulate changes in the operating state of the power grid on multiple time scales, and relies on real-person instructors for training, resulting in low usage and poor practicality.
A virtual teacher training system for grid dispatching is designed, including a virtual teacher designer, controller, actuator and management center. By setting the experience parameters and operating rules of the virtual teacher, a long process operation section and initial expected fault set are generated. Conduct probability failure calculation and evaluation to realize automated training lesson plan generation and evaluation.
Improve the intelligence and practicality of scheduling training, realize a comprehensive digital replacement of the role of real-person instructors, provide one-to-many, 24-hour and personalized scheduling training simulation services, enhance students' scheduling skills and coping with complex failures. ability.
Smart Images

Figure CN118967394A_ABST
Abstract
Description
Technical Field
[0001] The invention provides a method and system for generating a teaching plan for training a virtual instructor for power grid dispatching, and belongs to the technical field of power grid simulation. Background Art
[0002] New power systems are characterized by increased uncertainty in their operating states, an increase in vulnerable components, and complex fault patterns, paths, and characteristics. The diverse changes in grid operation modes and the variety of grid fault types are placing higher demands on the coordinated fault handling and control and operation personnel of multi-level grids. Conducting joint dispatching exercises is an effective means of improving the coordinated fault handling capabilities of control and operation personnel. Currently, the dispatching profession primarily relies on dispatcher training simulations for joint anti-accident exercises and personnel training. However, these techniques still lack the necessary simulation and interactive methods, resulting in low utilization and poor practicality of the exercise systems.
[0003] First, current training simulations are based on a single fixed grid operation mode for fault simulation and calculation, which is disconnected from the actual grid operation state and cannot simulate the changes in the grid operation state on multiple time scales; Second, real instructors are required to prepare lesson plans before training, monitor and control the drill process during training, and conduct comprehensive evaluations after training. This requires high professional skills of instructors and makes it difficult to allocate real instructors to support dispatcher training and drills on time and as needed. Third, in terms of interaction and progress during the exercise, the exercise was carried out step by step according to the script prepared in advance, with the problem of focusing more on "performance" and less on "practice", which failed to reflect the control personnel's ability to deal with sudden failures. Summary of the Invention
[0004] The present invention proposes a method and system for generating teaching plans for virtual instructor training of power grid dispatching. The present invention can provide dispatching training simulation services, thereby improving the intelligence and practicality of dispatching training simulation.
[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: In a first aspect, the present invention provides a virtual instructor training system for power grid dispatching, comprising: A virtual instructor designer is used to set the experience parameters and operating rules of the virtual instructor during the simulation training process; the operating rules include simulation operation rules, simulation fault triggering rules and simulation evaluation rules; A virtual instructor controller, the virtual instructor controller comprising: A module for generating lesson plans for initial training of virtual trainers, used to generate long-term operation sections of the power grid and initial anticipated fault sets; The virtual instructor probabilistic fault trigger module is used to calculate probabilistic faults, problem difficulty, and student skill matching based on equipment operating power and abnormal operating time, and automatically generate training lesson plans that include long-term grid operation modes and probabilistic faults. The virtual instructor training evaluation module is used to evaluate and score the trainees' performance during the entire training process after the simulation is completed; The virtual trainer executor is used to control the entire training process according to the calculation logic of the virtual trainer controller; And the virtual teacher management center is used to collect the control instructions of the virtual teacher controller and the operating status of the virtual teacher actuator in real time.
[0006] As a further improvement of the present invention, in the virtual instructor designer, the experience parameters and operating rules of the virtual instructor in the simulation training process are set by an experienced dispatcher based on historical training conditions, and the control centers at all levels specify different rules based on their respective power grid characteristics and the overall skill level of their personnel.
[0007] As a further improvement of the present invention, the simulation operation rules include: When the training time reaches the total simulation time, the virtual instructor automatically stops the training simulation. The total simulation time is the product of the maximum number of simulated faults and the upper limit of the single fault handling time. The virtual instructor loads the long-term operation section of the power grid according to the simulation time interval of the long-term operation section of the power grid; The virtual instructor needs to determine whether to continue the simulation training based on the simulation end conditions.
[0008] As a further improvement of the present invention, the simulation end condition includes: a) The simulation duration is greater than the total simulation duration; b) The current fault handling time is greater than the upper limit of single fault handling time; c) The number of triggered probabilistic grid faults has reached the maximum number of simulated faults, and the student has resolved the last grid fault; d) The power flow calculation does not converge.
[0009] As a further improvement of the present invention, the simulation fault triggering rule includes: Set the number of initial expected faults and the initial expected fault triggering time interval, and the virtual instructor will automatically trigger the initial expected fault according to the initial expected fault triggering time interval; The virtual instructor calls the probability fault calculation module once every time interval of the minimum triggering time interval of multiple probability faults; The virtual instructor evaluates the rapidity of fault handling based on the standard handling time for a single fault and the upper limit handling time for a single fault; The total duration of this training simulation is limited by setting the maximum number of simulated faults. When the number of faults triggered during this training simulation reaches the maximum number of simulated faults, probabilistic fault calculation and triggering will no longer be performed. According to the average score, the difficulty matching range of the questions is set to obtain the matching rules between the difficulty of the questions and the level of the students.
[0010] As a further improvement of the present invention, the simulation evaluation rules include: The maximum value of active power flow transfer entropy and voltage transfer entropy are set according to the operation characteristics of each power grid and historical operation experience; The active power over-limit coefficient and voltage over-limit coefficient are obtained according to the influence of the active power over-limit branch and the voltage over-limit node on the operation of the power grid; The normalized weight coefficient of the problem difficulty is calculated based on the weight ratio of the problem among the difficulty of active power flow adjustment, voltage regulation and multi-agency dispatch coordination; The fault handling correctness score distribution ratio is obtained based on the weight ratio of the equipment over-limit handling score, the power supply restoration score of the power outage equipment, and the fault point isolation score in the fault handling correctness score; The individual deduction points for the rationality of fault handling are calculated based on the deduction points for a single over-limit event, a single power outage device, or a single misoperation behavior. The normalized weight coefficient of the comprehensive score is obtained according to the weight ratio among the correctness, rationality and speed of fault handling.
[0011] As a further improvement of the present invention, the virtual instructor initial training lesson plan generation module includes: The long-term operation section generation unit of the power grid is used to: Obtain the current state estimation section from the control cloud platform as the initial simulation section; According to the total duration of this simulation , obtain the daily power generation plan, new energy forecast and load forecast data in the future simulation period, and form m future power grid operation sections, where , l is the time granularity of intraday real-time plan forecast; According to the simulation time interval of the long process operation section of the power grid , perform linear interpolation on m future power grid operation sections to obtain k long-term power grid operation sections, where ; Verify the power flow convergence of k long-term operation sections of the power grid; The initial expected fault set generation unit is used to: According to the initial expected number of faults and initial expected fault triggering time interval To formulate the initial set of expected faults; if , then conduct N-1 static safety analysis on the initial operation section and select the equipment failure with the greatest risk in grid operation as the initial expected failure; if , then N-1 static safety analysis is performed on the initial simulation section and the rth long-term operation section of the power grid, and the equipment failure with the greatest power grid operation risk under each section is selected as the initial expected fault set; among them, .
[0012] As a further improvement of the present invention, in the virtual instructor probability fault trigger module, Perform probabilistic failure calculations, the probability of equipment failure for
[0013] in, is the failure probability of equipment i at time t without considering historical operating conditions; is the influencing factor, which indicates the degree of influence of equipment load rate and time on the growth of failure probability. The abnormal operation time of the equipment. The maximum allowed abnormal operation time of the equipment.
[0014] As a further improvement of the present invention, the virtual instructor probability fault trigger module performs question difficulty calculation, and the question difficulty D is:
[0015] Where, The difficulty of the question, is the normalized power grid active power flow transfer entropy weight coefficient, is the normalized grid voltage transfer entropy weight coefficient, is the fault handling management difficulty weight coefficient, ; is the normalized power grid active power flow transfer entropy, is the normalized grid voltage transfer entropy, S Increase the difficulty of troubleshooting management.
[0016] As a further improvement of the present invention, the virtual instructor training evaluation module includes: The treatment rationality scoring unit is used to calculate the treatment rationality score during the entire simulation process. The student's operation does not further cause new equipment to exceed the limit or power outage, and the operation complies with the five-prevention operation regulations of the power grid. ; The handling speed score unit is used to require that fault handling be completed within the specified time. The longer the completion time, the lower the handling speed score. If the maximum fault handling time is exceeded, the handling speed score will be is 0; The comprehensive score unit is used to indicate that if the score of the handling correctness or handling speed is 0, it indicates that the trainee does not have the ability to handle the fault and the comprehensive score is 0; in other cases, the comprehensive score is the normalized calculated value of the above three scores.
[0017] As a further improvement of the present invention, the virtual instructor executor is specifically used to feed back the grid operation status and trainee behavior to the virtual instructor controller in real time, call the corresponding modules in the virtual instructor controller for calculation according to the training process, and receive control instructions to guide the simulation process into different branches.
[0018] As a further improvement of the present invention, the virtual instructor management center includes: The training lesson plan management unit is used to store the long-term operation section of the power grid, expected faults and probabilistic faults in the training simulation into the lesson plan library, and use the difficulty of the questions as the difficulty label of the lesson plan; The trainee skill management unit is used to store the trainee operations in the training simulation into the trainee behavior library and store the trainee scores in this training simulation into the trainee portrait library; The operation rule optimization unit is used to adjust the relevant parameters and rules in the virtual instructor designer according to the difficulty distribution of the lesson plans in the lesson plan library and the skill level distribution of the students.
[0019] In a second aspect, the present invention provides a method for generating a teaching plan for a virtual instructor training for power grid dispatching, comprising: Obtain relevant data on power grid regulation and control, and call upon the virtual instructor's experience parameters and operating rules during simulation training; The initial power flow section of the power grid simulation is obtained based on the acquired power grid control related data; Based on the initial power flow section of the power grid simulation and the experience parameters and operation rules of the virtual instructor during the simulation training process, the long-term operation section of the power grid and the initial expected fault set are generated; Based on the long-term operation section of the power grid and the initial expected fault set, and based on the equipment operating power and abnormal operation time, probabilistic fault calculations, question difficulty calculations, and student skill matching are performed to automatically generate training lesson plans that include the long-term operation mode of the power grid and probabilistic faults.
[0020] As a further improvement of the present invention, the generating of the long-term operation section of the power grid and the initial expected fault set includes: Obtain the current state estimation section from the control cloud platform as the initial simulation section; According to the total duration of this simulation , obtain the daily power generation plan, new energy forecast and load forecast data in the future simulation period, and form m future power grid operation sections, where , l is the time granularity of intraday real-time plan forecast; According to the simulation time interval of the long process operation section of the power grid , perform linear interpolation on m future power grid operation sections to obtain k long-term power grid operation sections, where ; Verify the power flow convergence of k long-term operation sections of the power grid; According to the initial expected number of faults and initial expected fault triggering time interval To formulate the initial set of expected faults; if , then conduct N-1 static safety analysis on the initial operation section and select the equipment failure with the greatest risk in grid operation as the initial expected failure; if , then N-1 static safety analysis is performed on the initial simulation section and the rth long-term operation section of the power grid, and the equipment failure with the greatest power grid operation risk under each section is selected as the initial expected fault set; among them, .
[0021] As a further improvement of the present invention, in the probability failure calculation, the failure probability of the equipment for:
[0022] in, is the failure probability of equipment i at time t without considering historical operating conditions; is the influencing factor, which indicates the degree of influence of equipment load rate and time on the growth of failure probability. The abnormal operation time of the equipment. The maximum allowed abnormal operation time of the equipment.
[0023] As a further improvement of the present invention, in the problem difficulty calculation, the problem difficulty D is:
[0024] Where, is the normalized power grid active power flow transfer entropy weight coefficient, is the normalized grid voltage transfer entropy weight coefficient, is the fault handling management difficulty weight coefficient, ; is the normalized power grid active power flow transfer entropy, is the normalized grid voltage transfer entropy, and S is the difficulty of fault handling management.
[0025] As a further improvement of the present invention, the method of automatically generating a training lesson plan including the long-term operation mode of the power grid and the probabilistic faults based on the long-term operation section of the power grid and the initial expected fault set, and based on the operating power of the equipment and the abnormal operation time, performing probabilistic fault calculation, question difficulty calculation, and student skill matching, includes: S401, loading the initial grid operation section; S402, constructing an initial fault set and performing initial fault simulation based on the initial power grid operation section; S403, performing power flow calculation based on the initial fault simulation to obtain an initial power flow section of the power grid simulation; S404: During the simulation clock advance period, determine whether the simulation exit condition is reached: If yes, S405, evaluate the simulation results and end the simulation; If not, S406, determine whether the power grid long process operation section simulation time interval has been reached: if so, S407, load the power grid long process simulation section, and then proceed to S408; if not, directly enter S408; S408: Determine whether the initial expected fault triggering time interval has been reached: If yes, S409, perform initial expected fault simulation and proceed to S410; if no, directly proceed to S410; S410, determining whether the minimum triggering time interval of multiple probability faults has been reached: If not, return to S403; If yes, S411, determine whether the maximum number of simulated faults has been reached: If yes, S412, return to S403; If not, S413, sample the current probability fault set, calculate the difficulty of the question, and determine whether it matches the student's skills based on the difficulty of the question; If yes, S414 simulates a probabilistic fault; returns to S403 and performs power flow calculation; If not, return to S404.
[0026] As a further improvement of the present invention, in S405, performing simulation result evaluation includes: The correctness score of the treatment is evaluated by judging the final operating state of the power grid at the end of the simulation , including equipment out-of-limit disposal points , Power supply restoration score for power outage equipment and fault point isolation score ; Calculate the rationality score of the operation during the entire simulation process, if the student's operation does not cause further equipment limit violations or power outages, and the operation complies with the five-prevention operation regulations of the power grid. ; Calculating the speed score :
[0027] in, The actual time taken by students to handle the faults. is the standard processing time for a single fault, is the upper limit of single fault handling time, and Set up in the Virtual Instructor Designer; Get a comprehensive score. If the score of handling correctness or handling speed is 0, it means that the trainee does not have the ability to handle the fault, and the comprehensive score is 0. In other cases, the comprehensive score is the normalized calculated value of the above three scores. The simulation results are evaluated based on the comprehensive scores.
[0028] In a third aspect, the present invention provides a system for generating lesson plans for training virtual instructors for power grid dispatching, comprising: The acquisition and calling module is used to obtain data related to power grid regulation and call the experience parameters and operating rules of the virtual instructor during the simulation training process; The initial power flow section calculation module is used to obtain the initial power flow section of the power grid simulation based on the acquired power grid control related data; The operation section generation module is used to generate the long-term operation section of the power grid and the initial expected fault set based on the initial power flow section of the power grid simulation and the experience parameters and operation rules of the virtual instructor during the simulation training process; The training lesson plan generation module is used to automatically generate training lesson plans that include the long-term operation mode of the power grid and probabilistic faults based on the long-term operation section of the power grid and the initial expected fault set, as well as the probabilistic fault calculation, question difficulty calculation, and student skill matching based on the equipment operating power and abnormal operation time.
[0029] In a fourth aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for generating lesson plans for training virtual instructors for power grid dispatching when executing the computer program.
[0030] In a fifth aspect, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method for generating lesson plans for training virtual instructors for power grid dispatching.
[0031] In a sixth aspect, the present invention provides a computer program product, which includes computer instructions, and the computer instructions instruct a computer to execute the method for generating a lesson plan for training a virtual instructor for power grid dispatching.
[0032] The beneficial effects of the present invention compared to the prior art are: The present invention proposes a method for constructing a virtual instructor for power grid dispatching. Through a virtual instructor designer, a virtual instructor controller, a virtual instructor executor and a virtual instructor management center, the power grid operation mode is transformed from a single-section operation mode to a time-series advancement of the power grid long-process operation mode. In terms of fault simulation, the method is transformed from deterministic expected fault simulation to probabilistic fault simulation related to the real-time operation mode. In terms of simulation interaction, the method is transformed from students manually inputting training scenarios and question difficulty to automatically matching questions according to students' skill levels, thereby realizing a comprehensive digital replacement of the role of real instructors in dispatching skills training, realizing a one-to-many, 24-hour all-weather and personalized dispatching training simulation service, and improving the intelligence and practicality of dispatching training simulation. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following introduction is made to the drawings of the embodiments of the present invention or the related technical solutions in the prior art. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative work.
[0034] Figure 1 A schematic diagram of the overall architecture of a virtual instructor for power grid dispatching according to an embodiment of the present invention; Figure 2 A schematic diagram of a virtual teacher rule design system according to an embodiment of the present invention; Figure 3 This is a functional diagram of a virtual instructor controller for power grid dispatching according to an embodiment of the present invention; Figure 4 A schematic diagram of the operation flow of the power grid dispatching virtual instructor executor according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the operation branches of the power grid dispatching virtual teacher executor according to an embodiment of the present invention. DETAILED DESCRIPTION
[0035] The embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention and are not to be construed as limiting the present invention. The step numbers in the following embodiments are provided for ease of explanation only and do not limit the order of the steps. The order of execution of the steps in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.
[0036] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, and connecting should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.
[0037] Knowledge and skills training in power system-related fields is an important part of enterprise talent development. Currently, a large number of simulation systems and training resources put into operation have certain problems such as difficulty in use, low resource utilization, and low training efficiency. The method for constructing a virtual instructor for power grid dispatching proposed in this invention can provide a technical foundation and R&D framework for the digital and intelligent development of other professional knowledge and skills training exercises.
[0038] This paper proposes a method for generating training lesson plans based on a virtual instructor for power grid dispatching. This method aims to completely replace the role of real instructors in dispatching skills training through digital means, thereby realizing intelligent, personalized, and all-weather dispatching training simulation services. This includes the following aspects: Virtual Instructor Designer: Sets the experience parameters and operating rules for the virtual instructor in simulation training. Experienced dispatchers can manually set these parameters based on historical training data. Control centers at all levels can specify different rules based on grid characteristics and personnel skill levels. The rule design system includes simulation operation rules, simulation fault triggering rules, and simulation evaluation rules.
[0039] The first object of the present invention is to provide a virtual instructor training system for power grid dispatching, comprising: A virtual instructor designer is used to set the experience parameters and operating rules of the virtual instructor during the simulation training process; the operating rules include simulation operation rules, simulation fault triggering rules and simulation evaluation rules; A virtual instructor controller, the virtual instructor controller comprising: A module for generating lesson plans for initial training of virtual trainers, used to generate long-term operation sections of the power grid and initial anticipated fault sets; The virtual instructor probabilistic fault trigger module is used to calculate probabilistic faults, problem difficulty, and student skill matching based on equipment operating power and abnormal operating time, and automatically generate training lesson plans that include long-term grid operation modes and probabilistic faults. The virtual instructor training evaluation module is used to evaluate and score the trainees' performance during the entire training process after the simulation is completed; The virtual trainer executor is used to control the entire training process according to the calculation logic of the virtual trainer controller; And the virtual teacher management center is used to collect the control instructions of the virtual teacher controller and the operating status of the virtual teacher actuator in real time.
[0040] The present invention first generates an initial training lesson plan and triggers probabilistic faults. The initial training lesson plan includes a long-term grid operation section and an initial set of anticipated faults. Fault probabilities are calculated based on equipment operating power and abnormal operating time, and the difficulty of questions is matched to the trainee's skill level. The virtual trainer actuator controls the training process according to the computational logic of the virtual trainer controller. The controller is fed back real-time grid status and trainee behavior. The simulation process is directed into different branches based on control instructions. Simulation execution of initial and probabilistic faults is supported. Control instructions and actuator status are aggregated in real time, providing training lesson plan management, trainee skills, and operating rule optimization. This enables centralized management of training lesson plans. Continuous optimization of trainee skill assessment and operating rules is achieved. Simulation realism is improved by transitioning from a single-section operation mode to a long-term operation mode. Probabilistic faults are simulated in conjunction with real-time operation mode to enhance the diversity and randomness of fault simulations. Question difficulty is automatically matched to the trainee's skill level, achieving personalized training. Simulation evaluation rules are used to comprehensively assess the trainee's fault handling capabilities, providing real-time feedback on training effectiveness.
[0041] Therefore, the present invention improves trainees' dispatching skills and ability to cope with complex power grid faults through personalized and intelligent training methods.
[0042] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0043] The present invention combines the concept of virtual digital human and robotic process automation technology to propose a method for constructing a virtual instructor for power grid dispatching. Through a virtual instructor designer, a virtual instructor controller, a virtual instructor executor and a virtual instructor management center, the power grid operation mode is transformed from a single-section operation mode to a time-series advancement of a long-process operation mode of the power grid; the fault simulation is transformed from deterministic expected fault simulation to probabilistic fault simulation related to the real-time operation mode; and the simulation interaction is transformed from students manually inputting training scenarios and question difficulty to automatically matching questions according to students' skill levels, thereby realizing a comprehensive digital replacement of the role of real instructors in dispatching skills training, and realizing a one-to-many, 24-hour all-weather and personalized dispatching training simulation service, thereby improving the intelligence and practicality of dispatching training simulation.
[0044] The overall architecture of the power grid dispatching virtual instructor training system based on the power grid dispatching virtual instructor is as follows: Figure 1 As shown below. Figure 1 Provide detailed description of each part.
[0045] 1. Virtual Instructor Designer: The virtual instructor designer is used to set the experience parameters and operating rules of the virtual instructor during the simulation training process. It can be manually set by experienced dispatchers based on historical training conditions. Control centers at all levels can specify different rules based on their respective power grid characteristics and overall personnel skill levels. The rule design system for virtual instructors is as follows: Figure 2 shown.
[0046] (1) Simulation operation rules 1) Total simulation time : It is the product of the maximum number of simulated faults and the upper limit of single fault handling time. When the training time reaches the total simulation time, the virtual instructor automatically stops the training simulation.
[0047] 2) Simulation time interval of long-term operation section of power grid ; Virtual teacher follows The time interval is used to load the long process operation section of the power grid.
[0048] 3) Simulation End Conditions: The virtual instructor needs to determine whether to continue the simulation training based on factors such as the student's skill level, fault handling time, and simulation progress. When the following rules are met, indicating that the student's skill level is insufficient or has reached the training goal, no further probabilistic faults will be triggered, and the virtual instructor will automatically stop the training simulation: a. The simulation duration is longer than the total simulation duration; b. The current fault handling time is greater than the upper limit of single fault handling time; c. The number of triggered probabilistic grid faults has reached the maximum number of simulated faults, and the student has resolved the last grid fault; d. The power flow calculation does not converge.
[0049] (2) Simulation fault triggering rules 1) Initial expected number of faults : The triggering of probabilistic faults is random. In order to ensure that expected faults are triggered during this training, the initial number of expected faults needs to be set. The minimum value is 1, and the maximum value can be determined according to the student's skill level. The higher the student's skill level, The bigger.
[0050] 2) Initial expected fault triggering time interval :If the number of initial expected faults If it is greater than 1, you need to set the triggering time interval between multiple initial expected faults. The virtual instructor is based on Automatically trigger the initial expected fault.
[0051] 3) Minimum triggering time interval for multiple probability faults :Probabilistic faults cannot be triggered frequently and irregularly. The virtual teacher Time, call the probability fault calculation module once.
[0052] 4) Standard handling time for single fault and the upper limit of single fault handling time :It is related to the difficulty of troubleshooting. The virtual instructor and Conduct rapid fault handling assessment.
[0053] 5) Maximum number of simulated faults :It is related to the skill level of the trainees. The total duration of the training simulation is limited by setting the maximum number of simulated faults. When the number of faults triggered during the training simulation reaches , the virtual instructor no longer performs probability fault calculation and triggering.
[0054] 6) Rules for matching question difficulty with student proficiency: The rules for matching question difficulty with student proficiency are shown in Table 1 below.
[0055] Table 1
[0056] (3) Simulation evaluation rules 1) Maximum value of active power flow transfer entropy and the maximum value of voltage transfer entropy : In order to normalize the active power flow transfer entropy and voltage transfer entropy, these two values are set according to the operation characteristics and historical operation experience of each provincial power grid.
[0057] 2) Active power over-limit coefficient and voltage over-limit coefficient : and It is to highlight the impact of active power exceeding limit branches and voltage exceeding limit nodes on grid operation, both of which are greater than 1.
[0058] 3) Normalized weight coefficient of problem difficulty: the weight ratio of the problem among the difficulty of active power flow adjustment, voltage regulation and multi-agency dispatch and coordination.
[0059] 4) Fault handling correctness score distribution ratio: Equipment over-limit handling score , Power supply restoration score for power outage equipment and fault point isolation score The weight ratio of the three aspects in the fault handling correctness score.
[0060] 5) Individual deduction points for rationality of fault handling: deduction points for a single limit-crossing event, a single power-off device, or a single misoperation.
[0061] 6) Normalized weight coefficient of comprehensive score: the weight ratio between the correctness of fault handling, the rationality of fault handling, and the speed of fault handling.
[0062] 2. Virtual Instructor Controller (1) Virtual instructor initial training lesson plan generation module: The initial training lesson plan includes two parts: the long-term operation section of the power grid and the initial expected fault set.
[0063] 1) Generation of long-term operation sections of power grid, including: Obtain the current state estimation section from the control cloud platform as the initial simulation section; According to the total duration of this simulation , obtain the daily power generation plan, new energy forecast and load forecast data in the future simulation period (the time granularity of the daily real-time plan forecast is l, for example, l is 15 minutes), and form m future power grid operation sections, where .
[0064] According to the simulation time interval of the long process operation section of the power grid , perform linear interpolation on m future power grid operation sections to obtain k long-term power grid operation sections, where The power flow convergence is verified for k long-term operation sections of the power grid. If the power flow does not converge, in order to ensure that the impact on the operation status of the power grid within the network is small, the power flow imbalance is allocated and adjusted by using equivalent units or equivalent loads at the boundary of the tie line to ensure the convergence of the power flow of the entire power grid.
[0065] 2) Generate the initial expected fault set, including: According to the initial expected number of faults and initial expected fault triggering time interval To formulate an initial set of anticipated faults.
[0066] if , then perform N-1 static safety analysis on the initial operation section and select the equipment failure with the greatest grid operation risk as the initial expected failure.
[0067] if , then N-1 static safety analysis is performed on the initial simulation section and the rth long-process operation section of the power grid, and the equipment fault with the greatest power grid operation risk under each section is selected as the initial expected fault set.
[0068] in,
[0069] (2) Virtual instructor probabilistic fault trigger module, including: The virtual instructor's probabilistic fault triggering module includes three parts: probabilistic fault calculation, question difficulty calculation, and student skill matching.
[0070] 1) Probabilistic failure calculation, including: According to operational reliability theory, the equipment outage probability p is closely related to the equipment's operating power. Linear segmentation is often used to fit the equipment outage probability curve that varies with power flow. This invention does not consider the probability of hidden protection failures. The formula is simplified as follows.
[0071]
[0072] Where, is the failure probability of equipment i at time t without considering historical operating conditions, is the rated power of device i, is the maximum operating power of device i.
[0073] Since the longer the equipment is overloaded or exceeds the limit, the greater the probability of equipment failure. If the trainees fail to deal with it in time, it will cause a wider range of failures. Therefore, the above failure probability function is superimposed with an exponentially growing time function.
[0074]
[0075] Where, is the failure probability of equipment i at time t considering historical operating conditions, is the influencing factor, which indicates the degree of influence of equipment load rate and time on the growth of failure probability. The abnormal operation time of the equipment (i.e. the operation time when the power is between the rated power and the limit operation power). This is the moment when the device first exceeds the rated power. The following constraints should be met
[0076] Where, The maximum allowed abnormal operation time of the equipment. , then the probability of equipment failure is 1.
[0077] After normalization of the above formula, the final equipment failure probability is obtained: for
[0078] Where, is the normalized failure probability of device i at time t.
[0079] The present invention mainly calculates the fault probability of four types of equipment, namely AC lines, transformers, busbars and converters, and obtains a probabilistic fault set through sampling.
[0080] 2) Question difficulty calculation, including: Each probabilistic fault triggers a new training question for the trainee. The difficulty of the question, that is, the difficulty of handling the fault, is calculated to provide a basis for subsequent assessment of the trainee's skill level. The difficulty of the question is assessed from both the perspective of grid operation and fault handling.
[0081] In terms of power grid operation, the power flow transfer entropy is used to describe the impact of power flow transfer in a short period of time on the entire power grid after a certain device is disconnected.
[0082] In the traditional power flow transfer entropy calculation method, the active power flow change distribution coefficient It represents the proportion of active power flow change borne by branch a after branch b is disconnected. The calculation formula is as follows
[0083] in, is the change in active power flow of branch a after branch b is disconnected, is the total number of power grid branches.
[0084] The traditional power flow transfer entropy calculation method does not consider the influence of branch limit transmission capacity on the rationality of power grid power distribution. The present invention adds active over-limit coefficient to the above calculation method. If the active power flow of branch a exceeds the limit after branch b is disconnected, the power flow change distribution coefficient of branch b should be appropriately expanded to indicate the irrationality of the current power grid power flow distribution.
[0085]
[0086] in, is the active power of branch a, is the upper limit of active power of branch a.
[0087] Power grid active power flow transfer entropy The calculation formula is as follows
[0088] Active power flow transfer entropy of power grid Perform normalization processing. If the active power flow of branch b is only borne by one branch, then The minimum value is 0. The maximum value of Perform normalization processing and set the maximum value of active power flow transfer entropy according to the characteristics of each provincial power grid Normalized power grid active power flow transfer entropy The calculation is as follows
[0089] Similarly, the grid voltage transfer entropy can be calculated and normalized grid voltage transfer entropy
[0090]
[0091]
[0092]
[0093] in, is the total number of grid buses, Represents the voltage change at node c after branch d is disconnected. It represents the voltage change ratio of node c after branch d is disconnected. is the voltage over-limit coefficient. If the voltage at node c exceeds the limit after branch d is disconnected, then It should be appropriately expanded to show the irrationality of the current grid voltage distribution, and the maximum value of voltage transfer entropy should be set according to the characteristics of each provincial grid. , is the voltage value of node c, is the lower limit of the voltage at node c, is the upper limit of the voltage at node c.
[0094] In terms of fault handling management, the number of control agencies and stations involved in the fault handling is counted based on the power-outage equipment, over-limit equipment, overloaded equipment and abnormal equipment affected by the fault. The larger the affected scope, the greater the scheduling difficulty, the longer the fault handling time, and thus the greater the difficulty of fault handling.
[0095]
[0096] Among them, S is the difficulty of fault handling management, D is the number of control agencies involved in this fault, and S is the number of plants and stations involved in this fault. is the total number of regulatory agencies, The total number of plants and stations.
[0097] Total number of regulatory agencies Calculation method: If the power grid at this level is nationally coordinated, the total number of regulatory agencies is the sum of the national, grid and provincial regulatory agencies, which is 34 in total.
[0098] If the power grid at this level is grid-regulated, the total number of regulating agencies is the sum of the national regulation, the grid regulation and all provincial regulation agencies dispatched by the grid regulation.
[0099] If the power grid at this level is provincially controlled, the total number of control agencies is the sum of the number of network control centers under the provincial control, the provincial control, and all local control centers under the provincial control.
[0100] If the power grid at this level is local regulation, the total number of regulatory agencies is the sum of the provincial regulation to which the local regulation belongs, the local regulation, and other local regulation centers that are topologically connected with the local regulation.
[0101] In summary, the calculation formula for the difficulty D of the question is:
[0102] Where, is the normalized power grid active power flow transfer entropy, is the normalized grid voltage transfer entropy, S is the difficulty of fault handling management, is the normalized power grid active power flow transfer entropy weight coefficient, is the normalized grid voltage transfer entropy weight coefficient, These three coefficients are set in the virtual instructor designer to meet the following constraints: .
[0103] 3) Student skills matching, including: The difficulty of the question and the matching rules of the student's skills specified in the virtual instructor designer are used to determine whether the current question should be pushed to the student.
[0104] (3) Virtual Instructor Training Evaluation Module After the simulation, the virtual instructor evaluates and scores the trainees' performance throughout the training process, evaluating them based on three aspects: correctness of handling, rationality of handling, and speed of handling.
[0105] 1) Correctness score of handling : By judging the final operating state of the power grid at the end of the simulation, the correctness of the fault handling is evaluated. It mainly includes the equipment over-limit handling score , Power supply restoration score for power outage equipment and fault point isolation score The total score for the correctness of the treatment is 100 points, and the distribution ratio of the three aspects is determined by the virtual teacher designer.
[0106] .
[0107] 2) Reasonableness score of disposal : The rationality of the handling requires that during the entire simulation process, the trainees' operations do not cause new equipment to exceed the limit or cause equipment power outages, and the operations comply with the five-prevention operation regulations of the power grid.
[0108]
[0109] in, is the number of devices exceeding the limit, Points deducted for a single limit-crossing event; is the number of power outage devices, Point deduction for individual power outage equipment; is the number of misoperated devices, Points are deducted for a single incorrect operation.
[0110] 3) Processing speed score : Speed of handling requires trainees to complete fault handling within the specified time. The longer the completion time, the lower the speed of handling score. If the maximum time for fault handling is exceeded, the speed of handling score will be 0.
[0111]
[0112] in, The actual time taken by students to handle the faults. is the standard processing time for a single fault, is the upper limit of single fault handling time, and Set up in the Virtual Trainer Designer.
[0113] 4) Comprehensive score: If the score for either the correctness or speed of handling is 0, it indicates that the trainee does not have the ability to handle the fault, and the overall score is 0. In other cases, the overall score is the normalized value of the above three scores.
[0114]
[0115] in, is the treatment accuracy weight coefficient, Disposal rationality weight coefficient, To handle the speed weight coefficient, these three coefficients are set in the virtual teacher designer 3. Virtual instructor actuator, including: The virtual instructor actuator controls the entire training process according to the calculation logic of the virtual instructor controller. Figure 4 During the virtual trainer's execution, the grid's operating status and trainee behavior are fed back to the virtual trainer controller in real time. Based on the training process, the corresponding modules in the virtual trainer controller are called to perform calculations and receive control instructions, guiding the simulation process into different branches.
[0116] The specific process includes the following steps: S401, loading the initial grid operation section; S402, constructing an initial fault set and performing initial fault simulation based on the initial power grid operation section; S403, performing power flow calculation based on the initial fault simulation to obtain an initial power flow section of the power grid simulation; S404, during the simulation clock advance period, determining whether a simulation exit condition is reached; If yes, S405, evaluate the simulation results and end the simulation; If not, S406, determining whether the power grid long process operation section simulation time interval has been reached; If yes, S407, load the long process simulation section of the power grid, and then proceed to S408; if no, directly enter S408; S408, determining whether the initial expected fault triggering time interval has been reached; If yes, S409, perform initial expected fault simulation and proceed to S410; if no, directly proceed to S410; S410, determining whether the minimum triggering time interval of multiple probability faults has been reached; If not, return to S403; If yes, S411, determine whether the maximum number of simulated faults has been reached; If yes, S412, return to S403; If not, S413, sample the current probability fault set, calculate the difficulty of the question, and determine whether it matches the student's skills based on the difficulty of the question; If yes, S414 simulates a probabilistic fault and returns to S403 to perform power flow calculation; If not, return to S404.
[0117] Figure 5 This is a schematic diagram of the operation branches of the virtual trainer actuator for power grid scheduling. The horizontal axis is the initial fault development process, and the vertical axis is the probabilistic fault development process during the simulation. Different trainees' operations will lead to different power grid operating states, triggering different probabilistic faults, and thus entering different simulation branches.
[0118] in, Indicates the initial fault development time, Indicates the probability of failure development time, t1, t2, t3 are three time points, is the initial fault, n is different time points, is the probability failure, n is 1, 23, 4….
[0119] 4. Virtual Teacher Management Center, including: The virtual instructor management center collects the control instructions of the virtual instructor controller and the operating status of the virtual instructor actuator in real time, and has three major functions: training lesson plan management, student skills and operating rule optimization.
[0120] 1) Training lesson plan management: The long-term operation section of the power grid, expected faults, and probabilistic faults simulated in this training are stored in the lesson plan database. The difficulty of the questions is used as the difficulty label of this lesson plan to facilitate subsequent analysis of the difficulty of the questions in the lesson plan database.
[0121] 2) Trainee skills management: Store the trainee operations in this training simulation in the trainee behavior library, and store the trainee scores in this training simulation in the trainee portrait library to facilitate subsequent analysis of the trainee's skill level.
[0122] 3) Operational rule optimization: Based on the difficulty distribution of lesson plans in the lesson plan library and the distribution of student skill levels, the relevant parameters and rules in the virtual instructor designer are adjusted to ensure a reasonable distribution of lesson plan difficulty and a reasonable match of student skill levels throughout the system.
[0123] By building a virtual instructor for power grid dispatching, training lesson plans that cover the long-term operation mode of the power grid and probabilistic faults are automatically generated. The training process is automatically guided according to the trainees' skill level, realizing a comprehensive digital replacement of the role of real instructors in dispatching skills training. It provides one-to-many, 24-hour, all-weather and personalized dispatching training simulation services, thereby improving the training efficiency of dispatching professionals.
[0124] The probabilistic fault generation technology in this invention considers the impact of equipment load rate and equipment fault abnormality time on fault probability, and proposes a probabilistic fault generation method in scheduling training simulation. It can adjust the training lesson plan in real time according to the performance of trainees and the operation status of the power grid during the training process, realizing the transformation from "script-based" training to "script-free training".
[0125] The method for calculating the difficulty of questions and the technology for matching students' skills are used to evaluate the difficulty of training questions from the perspectives of power grid operation and fault handling. The impact of branch line limit transmission capacity on the rationality of power grid flow distribution is considered. Improved flow transfer entropy and voltage transfer entropy are used to evaluate the technical difficulty of fault handling. The fault impact range is used to evaluate the difficulty of fault handling management. A normalized matching method for question difficulty and student skills is proposed, thereby achieving "teaching students in accordance with their aptitude" in dispatcher training.
[0126] The virtual instructor online intelligent guidance technology forms an overall architecture for virtual instructor online intelligent guidance by designing four functional modules and interaction mechanisms: virtual instructor designer, controller, actuator, and management center, providing a complete solution for the integration and application of virtual instructors in scheduling training simulation systems.
[0127] This method is implemented using the dispatch training simulation system of the control cloud platform. Pilot projects have been implemented and applied, demonstrating its effectiveness and practicality. The control cloud platform has been incorporated into the unified construction project for the energy internet and has broad application prospects. This method can effectively improve the efficiency of accident handling by control personnel, reduce accident risks, and enhance the safety and stability of the power grid, with significant economic and social benefits.
[0128] A second object of the present invention is to provide a method for generating a teaching plan for a virtual instructor training for power grid dispatching, comprising: Obtain relevant data on power grid regulation and control, and call upon the virtual instructor's experience parameters and operating rules during simulation training; The initial power flow section of the power grid simulation is obtained based on the acquired power grid control related data; Based on the initial power flow section of the power grid simulation and the experience parameters and operation rules of the virtual instructor during the simulation training process, the long-term operation section of the power grid and the initial expected fault set are generated; Based on the long-term operation section of the power grid and the initial expected fault set, and based on the equipment operating power and abnormal operation time, probabilistic fault calculations, question difficulty calculations, and student skill matching are performed to automatically generate training lesson plans that include the long-term operation mode of the power grid and probabilistic faults.
[0129] As an optional solution, the generation of a long-term operation section of the power grid and an initial anticipated fault set includes: Obtain the current state estimation section from the control cloud platform as the initial simulation section; According to the total duration of this simulation , obtain the daily power generation plan, new energy forecast and load forecast data in the future simulation period, and form m future power grid operation sections, where , l is the time granularity of intraday real-time plan forecast; According to the simulation time interval of the long process operation section of the power grid , perform linear interpolation on m future power grid operation sections to obtain k long-term power grid operation sections, where ; Verify the power flow convergence of k long-term operation sections of the power grid; According to the initial expected number of faults and initial expected fault triggering time interval To formulate the initial set of expected faults; if , then conduct N-1 static safety analysis on the initial operation section and select the equipment failure with the greatest risk in grid operation as the initial expected failure; if , then N-1 static safety analysis is performed on the initial simulation section and the rth long-term operation section of the power grid, and the equipment failure with the greatest power grid operation risk under each section is selected as the initial expected fault set; among them, .
[0130] As an optional solution, in the probability failure calculation, the failure probability of the equipment for:
[0131] in, is the failure probability of equipment i at time t without considering historical operating conditions; is the influencing factor, which indicates the degree of influence of equipment load rate and time on the growth of failure probability. The abnormal operation time of the equipment. The maximum allowed abnormal operation time of the equipment.
[0132] As an optional solution, in the problem difficulty calculation, the problem difficulty D is:
[0133] Where, is the normalized power grid active power flow transfer entropy weight coefficient, is the normalized grid voltage transfer entropy weight coefficient, is the fault handling management difficulty weight coefficient, ; is the normalized power grid active power flow transfer entropy, is the normalized grid voltage transfer entropy, and S is the difficulty of fault handling management.
[0134] As an optional solution, the method automatically generates a training lesson plan covering the long-term operation mode of the power grid and the probabilistic faults based on the long-term operation section of the power grid and the initial expected fault set, and on the basis of the equipment operating power and abnormal operation time, performing probabilistic fault calculation, question difficulty calculation, and student skill matching, including: S401, loading the initial grid operation section; S402, constructing an initial fault set and performing initial fault simulation based on the initial power grid operation section; S403, performing power flow calculation based on the initial fault simulation to obtain an initial power flow section of the power grid simulation; S404: During the simulation clock advance period, determine whether the simulation exit condition is reached: If yes, S405, evaluate the simulation results and end the simulation; If not, S406, determine whether the power grid long process operation section simulation time interval has been reached: if so, S407, load the power grid long process simulation section, and then proceed to S408; if not, directly enter S408; S408: Determine whether the initial expected fault triggering time interval has been reached: If yes, S409, perform initial expected fault simulation and proceed to S410; if no, directly proceed to S410; S410, determining whether the minimum triggering time interval of multiple probability faults has been reached: If not, return to S403; If yes, S411, determine whether the maximum number of simulated faults has been reached: If yes, S412, return to S403; If not, S413, sample the current probability fault set, calculate the difficulty of the question, and determine whether it matches the student's skills based on the difficulty of the question; If yes, S414 simulates a probabilistic fault; returns to S403 and performs power flow calculation; If not, return to S404.
[0135] As an optional solution, in S405 , performing simulation result evaluation includes: The correctness score of the treatment is evaluated by judging the final operating state of the power grid at the end of the simulation , including equipment out-of-limit disposal points , Power supply restoration score for power outage equipment and fault point isolation score ; Calculate the rationality score of the operation during the entire simulation process, if the student's operation does not cause further equipment limit violations or power outages, and the operation complies with the five-prevention operation regulations of the power grid. ; Calculating the speed score :
[0136] in, The actual time taken by students to handle the faults. is the standard processing time for a single fault, is the upper limit of single fault handling time, and Set up in the Virtual Instructor Designer; Get a comprehensive score. If the score of handling correctness or handling speed is 0, it means that the trainee does not have the ability to handle the fault, and the comprehensive score is 0. In other cases, the comprehensive score is the normalized calculated value of the above three scores. The simulation results are evaluated based on the comprehensive scores.
[0137] A third object of the present invention is to provide a system for generating lesson plans for training virtual instructors for power grid dispatching, comprising: The acquisition and calling module is used to obtain data related to power grid regulation and call the experience parameters and operating rules of the virtual instructor during the simulation training process; The initial power flow section calculation module is used to obtain the initial power flow section of the power grid simulation based on the acquired power grid control related data; The operation section generation module is used to generate the long-term operation section of the power grid and the initial expected fault set based on the initial power flow section of the power grid simulation and the experience parameters and operation rules of the virtual instructor during the simulation training process; The training lesson plan generation module is used to automatically generate training lesson plans that include the long-term operation mode of the power grid and probabilistic faults based on the long-term operation section of the power grid and the initial expected fault set, as well as the probabilistic fault calculation, question difficulty calculation, and student skill matching based on the equipment operating power and abnormal operation time.
[0138] The system for generating lesson plans for training virtual instructors for power grid dispatching of the present invention is based on a method for generating lesson plans for training virtual instructors for power grid dispatching.
[0139] The fourth object of an embodiment of the present invention is to provide an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned method for generating lesson plans for training virtual instructors for power grid dispatching when executing the computer program.
[0140] A fifth object of an embodiment of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned method for generating lesson plans for virtual instructor training for power grid dispatching.
[0141] A sixth object of an embodiment of the present invention is to provide a computer program product, wherein the computer program product includes computer instructions, and the computer instructions instruct a computer to execute the above-mentioned method for generating a lesson plan for a virtual instructor training for power grid dispatching.
[0142] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.
[0143] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.
[0144] The present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0145] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0146] Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.
[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the present invention.
Claims
1. A virtual instructor training system for power grid dispatching, It is characterized in that include: A virtual instructor designer is used to set the experience parameters and operation rules of the virtual instructor during the simulation training process; the operation rules include simulation operation rules, simulation fault triggering rules and simulation evaluation rules; A virtual instructor controller, the virtual instructor controller comprising: The module for generating lesson plans for initial training of virtual trainers is used to generate long-term operation sections of power grids and initial expected fault sets; The virtual instructor probabilistic fault trigger module is used to calculate probabilistic faults, calculate the difficulty of questions, and match trainee skills based on the equipment operating power and abnormal operating time, and automatically generate training lesson plans that include the long-term operation mode of the power grid and probabilistic faults. The virtual instructor training evaluation module is used for the virtual instructor to evaluate and score the trainees' performance during the entire training process after the simulation is completed; The virtual instructor executor is used to control the entire training process according to the calculation logic of the virtual instructor controller; The virtual teacher management center is used to collect the control instructions of the virtual teacher controller and the operating status of the virtual teacher actuator in real time.
2. A virtual instructor training system for power grid dispatching according to claim 1, It is characterized in that In the virtual instructor designer, the experience parameters and operating rules of the virtual instructor during the simulation training process are set by an experienced dispatcher based on historical training situations, and different levels of control centers specify different rules based on their respective power grid characteristics and the overall skill level of their personnel.
3. A virtual instructor training system for power grid dispatching according to claim 1, It is characterized in that The simulation operation rules include: When the training time reaches the total simulation time, the virtual instructor automatically stops the training simulation; the total simulation time is the product of the maximum number of simulated faults and the upper limit of the single fault handling time; The virtual instructor loads the long-process operation section of the power grid according to the simulation time interval of the long-process operation section of the power grid; The virtual instructor needs to determine whether to continue the simulation training based on the simulation end conditions.
4. A virtual instructor training system for power grid dispatching according to claim 3, It is characterized in that The simulation end condition includes: a) The simulation duration is greater than the total simulation duration; b) The current fault handling time is greater than the upper limit of single fault handling time; c) The number of triggered grid probabilistic faults has reached the maximum number of simulated faults, and the trainee has handled the last grid fault; d) The power flow calculation does not converge.
5. A virtual instructor training system for power grid dispatching according to claim 1, It is characterized in that The simulation fault triggering rule includes: Set the number of initial expected faults and the initial expected fault triggering time interval, and the virtual instructor automatically triggers the initial expected fault according to the initial expected fault triggering time interval; The virtual instructor calls the probability fault calculation module once every minimum triggering time interval of multiple probability faults; The virtual instructor evaluates the rapidity of fault handling based on the standard handling time for a single fault and the upper limit handling time for a single fault; The total duration of this training simulation is limited by setting the maximum number of simulated faults. When the number of faults triggered during this training simulation has reached the maximum number of simulated faults, the probability fault calculation and triggering will no longer be performed. According to the average score, the difficulty matching range of the questions is set to obtain the matching rules between the difficulty of the questions and the level of the students.
6. A virtual instructor training system for power grid dispatching according to claim 1, It is characterized in that The simulation evaluation rules include: The maximum value of active power flow transfer entropy and voltage transfer entropy are set according to the operation characteristics of each power grid and historical operation experience; The active over-limit coefficient and the voltage over-limit coefficient are obtained according to the influence of the active over-limit branch and the voltage over-limit node on the operation of the power grid; The normalized weight coefficient of the difficulty of the question is calculated according to the weight ratio between the difficulty of active power flow adjustment, the difficulty of voltage regulation and the difficulty of multi-agency dispatching and coordination; According to the weight ratio of the equipment over-limit handling score, the power supply restoration score of the power-off equipment and the fault point isolation score in the fault handling correctness score, the fault handling correctness score distribution ratio is obtained; The single deduction value for the rationality of fault handling is obtained based on the deduction value of a single over-limit event, a single power-off device, and a single misoperation behavior; The normalized weight coefficient of the comprehensive score is obtained according to the weight ratio among the correctness, rationality and speed of fault handling.
7. A virtual instructor training system for power grid dispatching according to claim 1, It is characterized in that The virtual instructor initial training lesson plan generation module includes: The long-term operation section generation unit of the power grid is used for: Obtain the current state estimation section from the control cloud platform as the initial simulation section; According to the total duration of this simulation , obtain the daily power generation plan, new energy forecast and load forecast data in the future simulation period, and form m future power grid operation sections, where , l is the time granularity of intraday real-time plan forecast; According to the long process of power grid operation section simulation time interval , perform linear interpolation on m future operation sections of the power grid to obtain k long-term operation sections of the power grid, where ; Verify the power flow convergence of k long-term operation sections of the power grid; The initial expected fault set generation unit is used to: According to the initial expected number of faults and initial expected fault triggering time interval To formulate an initial set of anticipated faults; if , then conduct N-1 static safety analysis on the initial operation section and select the equipment failure with the greatest risk of power grid operation as the initial expected failure; if , then N-1 static safety analysis is performed on the initial simulation section and the rth long-term operation section of the power grid, and the equipment failure with the greatest risk of power grid operation under each section is selected as the initial expected fault set; among them, .
8. A virtual instructor training system for power grid dispatching according to claim 1, It is characterized in that In the virtual instructor probability fault trigger module, Perform probabilistic failure calculations, the probability of equipment failure for in, is the failure probability of equipment i at time t without considering the historical operation conditions; is the influencing factor, which indicates the influence of equipment load rate and time on the growth of failure probability. The abnormal operation time of the equipment. The maximum allowed abnormal operation time of the equipment.
9. A virtual instructor training system for power grid dispatching according to claim 1, It is characterized in that In the virtual instructor probability fault trigger module, the question difficulty is calculated, and the question difficulty D is: In the formula, is the normalized power grid active power flow transfer entropy weight coefficient, is the normalized grid voltage transfer entropy weight coefficient, is the fault handling management difficulty weight coefficient, ; is the normalized power grid active power flow transfer entropy, is the normalized grid voltage transfer entropy, S Increase the difficulty of troubleshooting management.
10. A virtual instructor training system for power grid dispatching according to claim 1, It is characterized in that The virtual trainer training assessment module includes: The handling correctness scoring unit is used to evaluate the correctness of fault handling by judging the final operating state of the power grid at the end of the simulation. Including equipment out-of-limit handling score , Power supply restoration score for power outage equipment And the fault point isolation score ; The treatment rationality scoring unit is used to calculate the treatment rationality score during the entire simulation process. The trainee's operation does not further cause new equipment to exceed the limit or power outage, and the operation complies with the five-defense operation regulations of the power grid. ; The processing speed score unit is used to require that fault processing must be completed within a specified time. The longer the completion time, the lower the processing speed score. If the maximum fault processing time is exceeded, the processing speed score will be is 0; The comprehensive score unit is used to indicate that if the score of handling correctness or handling speed is 0, it indicates that the trainee does not have the ability to handle the fault and the comprehensive score is 0; in other cases, the comprehensive score is the normalized calculated value of the above three scores.
11. A virtual instructor training system for power grid dispatching according to claim 1, It is characterized in that The virtual instructor actuator is specifically used to feed back the grid operation status and trainee behavior to the virtual instructor controller in real time, call the corresponding modules in the virtual instructor controller for calculation according to the training process, and receive control instructions to guide the simulation process into different branches.
12. A virtual instructor training system for power grid dispatching according to claim 1, It is characterized in that The virtual instructor management center includes: The training lesson plan management unit is used to store the long-process operation section of the power grid, expected faults and probabilistic faults of the training simulation into the lesson plan library, and use the difficulty of the question as the difficulty label of the lesson plan; The trainee skill management unit is used to store the trainee operations in the training simulation into the trainee behavior library, and store the trainee scores in this training simulation into the trainee portrait library; The operation rule optimization unit is used to adjust the relevant parameters and rules in the virtual instructor designer according to the difficulty distribution of the lesson plans in the lesson plan library and the skill level distribution of the students.
13. A method for generating lesson plans for virtual instructor training in power grid dispatching. It is characterized in that include: Obtain relevant data on power grid regulation and control, and call upon the experience parameters and operating rules of the virtual instructor during simulation training; The initial power flow section of the power grid simulation is obtained according to the obtained power grid control related data; According to the initial power flow section of the power grid simulation and the experience parameters and operation rules of the virtual instructor in the simulation training process, the long-term operation section of the power grid and the initial expected fault set are generated; According to the long-term operation section of the power grid and the initial expected fault set, and based on the equipment operating power and abnormal operation time, probabilistic fault calculations, question difficulty calculations, and trainee skill matching are performed to generate a training lesson plan that includes the long-term operation mode of the power grid and probabilistic faults.
14. A method for generating teaching plans for virtual instructor training of power grid dispatching according to claim 13, It is characterized in that The generating of the long-term operation section of the power grid and the initial expected fault set includes: Obtain the current state estimation section from the control cloud platform as the initial simulation section; According to the total duration of this simulation , obtain the daily power generation plan, new energy forecast and load forecast data in the future simulation period, and form m future power grid operation sections, where , l is the time granularity of intraday real-time plan forecast; According to the long process of power grid operation section simulation time interval , perform linear interpolation on m future operation sections of the power grid to obtain k long-term operation sections of the power grid, where ; Verify the power flow convergence of k long-term operation sections of the power grid; According to the initial expected number of faults and initial expected fault triggering time interval To formulate an initial set of anticipated faults; if , then conduct N-1 static safety analysis on the initial operation section and select the equipment failure with the greatest risk of power grid operation as the initial expected failure; if , then N-1 static safety analysis is performed on the initial simulation section and the rth long-term operation section of the power grid, and the equipment failure with the greatest risk of power grid operation under each section is selected as the initial expected fault set; among them, .
15. A method for generating teaching plans for virtual instructor training of power grid dispatching according to claim 13, It is characterized in that In the above probability failure calculation, the failure probability of the equipment for in, is the failure probability of equipment i at time t without considering the historical operation conditions; is the influencing factor, which indicates the influence of equipment load rate and time on the growth of failure probability. The abnormal operation time of the equipment. The maximum allowed abnormal operation time of the equipment.
16. A method for generating teaching plans for virtual instructor training of power grid dispatching according to claim 13, It is characterized in that In the above-mentioned problem difficulty calculation, the problem difficulty D is: In the formula, is the normalized power grid active power flow transfer entropy weight coefficient, is the normalized grid voltage transfer entropy weight coefficient, is the fault handling management difficulty weight coefficient, ; is the normalized power grid active power flow transfer entropy, is the normalized grid voltage transfer entropy, and S is the difficulty of fault handling management.
17. A method for generating teaching plans for virtual instructor training for power grid dispatching according to claim 13, It is characterized in that According to the long-term operation section of the power grid and the initial expected fault set, the probability fault calculation, the question difficulty calculation, and the trainee skill matching are performed based on the equipment operating power and abnormal operation time, and a training lesson plan including the long-term operation mode of the power grid and the probability fault is automatically generated, including: S401, loading the initial power grid operation section; S402, constructing an initial fault set, and performing initial fault simulation based on an initial power grid operation section; S403, performing power flow calculation based on the initial fault simulation to obtain an initial power flow section of the power grid simulation; S404: During the simulation clock advancement period, determine whether the simulation exit condition is reached: If yes, S405, evaluate the simulation results and end the simulation; If not, S406, determine whether the power grid long process operation section simulation time interval has been reached: if so, S407, load the power grid long process simulation section, and then proceed to S408; if not, directly enter S408; S408, determining whether the initial expected fault triggering time interval has been reached: If yes, S409, perform initial expected fault simulation and proceed to S410; if no, directly proceed to S410; S410, determining whether the minimum triggering time interval of multiple probability faults has been reached: If not, return to S403; If yes, S411, determine whether the maximum number of simulated faults has been reached: If yes, S412, return to S403; If not, S413, sample the current probability fault set, calculate the difficulty of the question, and determine whether it matches the student's skills based on the difficulty of the question; If yes, S414, simulate probabilistic fault; return to S403, perform power flow calculation; If not, return to S404.
18. A method for generating teaching plans for virtual instructor training of power grid dispatching according to claim 17, It is characterized in that In S405, performing simulation result evaluation includes: The correctness score of the treatment is evaluated by judging the final operating state of the power grid at the end of the simulation , including equipment out-of-limit handling points , Power supply restoration score for power outage equipment And the fault point isolation score ; Calculate the rationality score of the operation during the entire simulation process, if the trainee's operation does not cause any further equipment over-limit or power outage, and the operation complies with the five-defense operation regulations of the power grid. ; Calculating the speed score : in, The actual time taken by students to handle the faults. is the standard processing time for a single fault, is the upper limit of single fault handling time, and Set up in the Virtual Instructor Designer; Get the comprehensive score. If the score of the correctness of handling or the speed of handling is 0, it means that the trainee does not have the ability to handle the fault, and the comprehensive score is 0. In other cases, the comprehensive score is the normalized calculated value of the above three scores. The simulation results are evaluated based on the comprehensive score.
19. A teaching plan generation system for virtual instructor training of power grid dispatching. It is characterized in that include: The acquisition and calling module is used to obtain the relevant data of power grid regulation and call the experience parameters and operation rules of the virtual instructor in the simulation training process; An initial power flow section calculation module is used to obtain the initial power flow section of the power grid simulation according to the acquired power grid control related data; The operation section generation module is used to generate the long-term operation section of the power grid and the initial expected fault set according to the initial power flow section of the power grid simulation and the experience parameters and operation rules of the virtual instructor in the simulation training process; The training lesson plan generation module is used to automatically generate training lesson plans that include the long-term operation mode of the power grid and probabilistic faults based on the long-term operation section of the power grid and the initial expected fault set, and based on the equipment operating power and abnormal operation time, perform probabilistic fault calculations, question difficulty calculations, and student skill matching.
20. An electronic device, It is characterized in that It includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for generating lesson plans for training virtual instructors for power grid dispatching as described in any one of claims 13 to 18 is implemented.
21. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for generating lesson plans for training virtual instructors for power grid dispatching as described in any one of claims 13-18 is implemented.
22. A computer program product comprising computer instructions, It is characterized in that The computer instructions instruct the computer to execute the method for generating teaching plans for virtual instructor training for power grid dispatching as described in any one of claims 13-18.
Citation Information
Patent Citations
Generation method and system of training lesson plan of dispatcher training simulation system
CN103050048A
Simulation lesson plan automatically-generating method for dispatcher training
CN103440811A
Power grid regulation and control operation simulation time sequence power flow simulation method, device, equipment and medium
CN115117887A
Generation method, system and equipment of power grid training teaching plan
CN116011946A
An implementation method of training system applied the power system restorative training algorithm
KR1020050041341A
