Health assessment-based master-dispatcher coordinated power distribution network operation decision method and system

Abstract

The application discloses a kind of based on health assessment main distribution collaborative power distribution network operation decision method and system, it is related to main distribution collaborative dispatching technical field.The application is first based on double-end multi-source operation data to construct main distribution collaborative health benchmark image and scoring rule, obtains current health assessment result containing collaborative health deviation distribution;Again, deviation distribution is dynamically optimized measurement configuration, and combined with health deviation weight, weighted state estimation and network parameter identification are carried out;Finally, based on accurate state estimation result and correction parameter, operation strategy is generated through pre-trained multi-agent collaborative decision model, and the model is continuously optimized through incremental learning.The application improves state estimation accuracy from data source and computer mechanism two dimensions, realizes the active control of power grid operation risk, guarantees the accuracy and reliability of main distribution collaborative decision.

Description

Technical Field

[0001] This invention relates to the field of main-distribution coordinated dispatching technology, and in particular to a method and system for making decisions on the operation of a main-distribution coordinated distribution network based on health assessment. Background Technology

[0002] With the large-scale integration of distributed power sources and flexible loads into the distribution network, the operational coupling between the main grid and the distribution network has significantly increased. Coordinated operation control between the main grid and distribution network has become a core technical means to ensure the safe and economical operation of the power grid. State estimation, as a key link in the coordinated operation state perception of the grid, can analyze core operational state quantities such as node voltage and branch power from measurement data. Its output accuracy directly determines the accuracy and reliability of coordinated decision-making between the main grid and distribution network.

[0003] Currently, the measurement systems of the main distribution network generally adopt the traditional configuration mode of fixed location and static planning. The installation location, data acquisition parameters and deployment density of the measurement equipment are all preset in advance, which cannot adapt to the dynamic fluctuation characteristics of the power grid and is difficult to match the differentiated sensing needs of different regions. There are generally problems of insufficient measurement coverage and missing key parameters in the main distribution interaction section, the section with severe fluctuations and the high-risk area.

[0004] Meanwhile, existing state estimation methods mostly use globally equal weights to process measurement data, ignoring the differences in operating states between different areas of the power grid and failing to compensate for the perception gap caused by insufficient measurement. The combined effect of measurement configuration deficiencies and state estimation mechanism defects leads to insufficient accuracy and distorted results in state estimation. This forces the power grid to only respond passively after operating parameters exceed limits, unable to identify and intervene in persistent deviations approaching risk in advance. This not only limits the selection of control measures but may also lead to errors in primary and secondary distribution coordination decisions due to distorted basic data, seriously affecting the reliability and operational safety of the power grid. Summary of the Invention

[0005] To address this issue, this invention provides a method and system for decision-making on the operation of a primary-distributor coordinated distribution network based on health assessment. This method solves the problem that the existing technology uses a static fixed mode for the configuration of primary-distributor coordinated network measurements and the weighting of state estimation, which is mismatched with the dynamic fluctuations in the operation characteristics of the power grid. This leads to insufficient accuracy and distorted results in state estimation, which in turn causes low accuracy and reliability of primary-distributor coordinated operation decisions.

[0006] To address the aforementioned technical problems, embodiments of the present invention provide a method for decision-making on the operation of a primary and secondary distribution network based on health assessment, the method comprising:

[0007] Based on the dual-end multi-source operation data of the main distribution coordinated power grid and the preset power grid health indicators, a health benchmark profile and health scoring rules for the main distribution coordinated power grid are constructed, and the current health assessment results including the distribution of coordinated health deviation are obtained by combining real-time measurement data.

[0008] Based on the collaborative health deviation distribution in the current health assessment results, high deviation spatial locations are screened and areas to be optimized are determined. The optimal measurement scheme is solved by combining candidate measurement points, and the measurement configuration optimization is completed to obtain the optimized measurement data.

[0009] Extract the collaborative health deviation distribution from the current health assessment results, set health deviation weights based on the spatial location of each measurement point after measurement configuration optimization, perform weighted state estimation based on the optimized measurement data and health deviation weights, and simultaneously use the state estimation residual sequence to identify parameters, thereby obtaining the state estimation results of the main and distribution coordinated power grid and the corrected network parameters.

[0010] Based on the state estimation results and the corrected network parameters, a short-term operating scenario is generated by combining the pre-generated multi-agent collaborative decision-making model, and collaborative decision-making is carried out to obtain a collaborative operating strategy that includes control instructions for each controllable device in the main and distribution networks.

[0011] Preferably, the construction of a health benchmark profile and health scoring rules for main-distribution coordination based on dual-end multi-source operation data of the main-distribution coordinated power grid and preset power grid health indicators includes:

[0012] Based on the dual-end multi-source operation data of the main grid and distribution grid, we extract the independent electrical quantity characteristics of the main grid side and the distribution grid side, as well as the coordinated characteristic quantities corresponding to the interaction status indicators of the main grid and distribution grid.

[0013] Based on historical operating data and the aforementioned collaborative characteristic quantities, the normal operating fluctuation range of each parameter on the main grid side, distribution grid side, and main-distribution interaction section and the correlation between parameters are determined, forming a health benchmark profile of main-distribution collaboration.

[0014] The pre-set power grid health indicators are coupled and weighted using the aforementioned collaborative characteristic quantities as adjustment factors to determine the deviation level threshold, dynamic weight coefficient, and weighted deduction calculation rules for each indicator, thereby forming a health scoring rule for main and distribution coordination.

[0015] Preferably, the step of obtaining the current health assessment result including the distribution of co-health deviations by combining real-time measurement data includes:

[0016] Real-time operating values ​​are extracted from the main grid side, distribution network side, and main-distribution interaction section based on real-time measurement data;

[0017] By comparing the real-time operating values ​​with the normal operating fluctuation range of the corresponding parameters in the health benchmark profile, the deviation of each power grid health indicator is calculated, and the deviation is corrected by combining the deviation of the correlation between parameters.

[0018] Based on the aforementioned health scoring rules, the corrected deviation is weighted and mapped to a level to obtain the current primary and secondary collaborative health score.

[0019] The deviation distribution of each power grid health indicator is statistically analyzed according to spatial region to generate a coordinated health deviation distribution, which is then combined with the current main and distribution coordinated health score to form the current health assessment result.

[0020] Preferably, the step of screening high-bias spatial locations and determining the area to be optimized based on the collaborative health deviation distribution in the current health assessment results, solving for the optimal measurement scheme by combining candidate measurement points, completing the measurement configuration optimization, and obtaining the optimized measurement data includes:

[0021] Spatial locations whose deviation exceeds a preset threshold are selected from the collaborative health deviation distribution, and the spatial locations and their electrical associated ranges are used as the areas to be optimized.

[0022] Based on the deviation of each spatial location within the area to be optimized, corresponding deviation weights are set. With the goal of minimizing the weighted state estimation error of the area to be optimized, and with the location and collectable parameters of the candidate measurement points as constraints, the optimal measurement scheme, which includes the installation location, type and quantity of measurement equipment, is obtained.

[0023] Based on the optimal measurement scheme, supplementary pseudo-measurement data is generated in the high-bias region, redundant measurement data is removed in the low-bias region, and the optimized measurement data is obtained by integration.

[0024] Preferably, the step of extracting the coordinated health deviation distribution from the current health assessment results, setting health deviation weights based on the spatial location of each measurement point after measurement configuration optimization, performing weighted state estimation based on the optimized measurement data and health deviation weights, and simultaneously using the state estimation residual sequence for parameter identification to obtain the state estimation results and corrected network parameters of the main and distribution coordinated power grid includes:

[0025] Based on the collaborative health deviation distribution corresponding to the spatial location of each measurement point, the health deviation weight of each measurement point is set by gradient, and a diagonal matrix of health deviation weights is constructed.

[0026] The optimized measurement data and the health deviation weighted diagonal matrix are input into the weighted least squares state estimation algorithm, and the state estimation results of voltage amplitude, phase angle and branch power flow of all nodes in the main distribution coordinated power grid are obtained by iterative solution.

[0027] The residual time series of each measurement point during the state estimation process is extracted. Combined with historical measurement data, the Kalman filter algorithm is used to identify the line impedance, transformer equivalent parameters, and main distribution interface coupling parameters to obtain the corrected network parameters.

[0028] Preferably, the pre-generation process of the multi-agent collaborative decision-making model is as follows:

[0029] Based on a pre-generated sample set of operating scenarios with health tags, the operating status characteristics and health assessment results of the main grid and distribution network are extracted and mapped to the state spaces of the main grid agent and each distribution network feeder agent, respectively.

[0030] Based on the type and physical regulation range of controllable equipment in the main distribution coordinated power grid, the discrete and continuous action spaces of each intelligent agent are defined.

[0031] Construct a multi-objective reward function that includes a line operation safety penalty, a network loss negative reward, and a health score change reward;

[0032] A multi-agent reinforcement learning algorithm is used to jointly train the main network agent and the agents of each distribution network feeder until the multi-objective reward function converges, thus obtaining a collaborative decision-making model.

[0033] Preferably, the process of generating the running scenario sample set is as follows:

[0034] Based on the load level, distributed power output level and network topology in historical operation data, and combined with the corresponding health assessment results, a clustering algorithm is used to divide the network into several typical operation categories.

[0035] Extract representative operational cross-sections of each category and add health labels to generate a sample set of basic operational modes;

[0036] Based on extreme event data, the basic samples are extended temporally to generate dynamic scene samples that include scenarios such as faults and sudden weather changes.

[0037] By integrating basic operation mode samples and dynamic scenario samples, an operation scenario sample set is obtained.

[0038] Preferably, the step of generating a short-term operating scenario and making collaborative decisions based on the state estimation results and the corrected network parameters, combined with a pre-generated multi-agent collaborative decision-making model, to obtain a collaborative operating strategy containing control commands for each controllable device in the main and distribution networks, includes:

[0039] Using the current state estimation result as the initial operating section, combined with the load forecast data and distributed power output forecast data for the future preset time period, the corrected network parameters are used to perform time-by-time power flow simulation to generate a short-term operating scenario sequence.

[0040] The short-term operation scenario sequence is input into the pre-generated collaborative decision-making model. The main grid agent outputs the main grid side voltage regulation and reactive power compensation control commands, and each distribution network feeder agent outputs the distribution network side reactive power regulation, energy storage charging and discharging and interruptible load control commands.

[0041] Integrate the control commands of all intelligent agents to form a master-slave collaborative operation strategy that satisfies timing constraints.

[0042] Preferably, after obtaining the collaborative operation strategy, the method further includes:

[0043] Collect actual measurement data after the collaborative operation strategy is executed, calculate the actual health score after execution according to the health scoring rules, and compare it with the health score before execution to obtain the change in health score.

[0044] Incremental samples are constructed based on the pre-execution running status, control instructions in the collaborative operation strategy, and changes in health scores after execution.

[0045] We employ a mini-batch gradient descent algorithm, combined with incremental samples, to incrementally learn and update the policy network and value network of the collaborative decision-making model.

[0046] This invention also provides a primary-distribution coordinated distribution network operation decision-making system based on health assessment. This system is used to implement the aforementioned primary-distribution coordinated distribution network operation decision-making method based on health assessment, specifically including:

[0047] The health assessment module is used to construct a health benchmark profile and health scoring rules for the main and distribution coordinated power grid based on the dual-end multi-source operation data of the main and distribution coordinated power grid and the preset power grid health indicators, and to obtain the current health assessment results including the distribution of coordinated health deviations by combining real-time measurement data.

[0048] The measurement optimization module is used to screen high-deviation spatial locations and determine the area to be optimized based on the collaborative health deviation distribution in the current health assessment results, solve the optimal measurement scheme by combining candidate measurement points, complete the measurement configuration optimization, and obtain the optimized measurement data.

[0049] The identification module is used to extract the collaborative health deviation distribution in the current health assessment results, set the health deviation weights based on the spatial location of each measurement point after the measurement configuration optimization, perform weighted state estimation based on the optimized measurement data and health deviation weights, and simultaneously use the state estimation residual sequence to identify parameters, thereby obtaining the state estimation results of the main and distribution coordinated power grid and the corrected network parameters.

[0050] The collaborative decision-making module, based on the state estimation results and the corrected network parameters, combines the pre-generated multi-agent collaborative decision-making model to generate short-term operating scenarios and make collaborative decisions, thereby obtaining a collaborative operating strategy that includes control instructions for each controllable device in the main and distribution networks.

[0051] As can be seen from the above technical solutions, this invention application has the following beneficial effects:

[0052] First, this invention quantifies the spatial distribution of power grid operation deviations through primary and secondary coordinated health assessment. It specifically supplements pseudo-measurements in high-deviation and high-risk areas and eliminates redundant measurements in low-risk areas. This solves the problems of traditional static measurement configuration being unable to adapt to the dynamic fluctuation characteristics of the power grid and insufficient measurement coverage in key areas. It also tilts measurement resources toward areas with higher sensing needs, providing higher-quality input data for state estimation.

[0053] Secondly, this invention sets measurement weights according to the gradient of health deviation in each region, which strengthens the contribution of high-risk region data in state calculation. At the same time, it uses the state estimation residual sequence to identify and correct the drift error of network parameters such as line impedance and transformer parameters online. This solves the problem that traditional equal weight state estimation easily ignores regional differences and offline parameter model error accumulation. It effectively avoids the continuous deviation being masked by averaging and significantly improves the reliability of main and distribution coordinated power grid state perception.

[0054] Third, this invention incorporates changes in health scores into the reward function of multi-agent reinforcement learning, enabling the decision-making model to not only focus on electrical quantity limit constraints but also proactively optimize the overall health status of the power grid, achieving a shift from "passive response after limit exceedance" to "proactive intervention during deviation periods." Simultaneously, by continuously updating the model through incremental learning after strategy execution, the adaptability of the decision to long-term operating condition changes such as equipment aging and load characteristic drift is improved, ensuring the accuracy and robustness of the main and distribution coordinated operation decision. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly described below. Referring to the drawings will make the features and advantages of the present invention clearer. The drawings are illustrative and should not be construed as limiting the present invention in any way. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0056] Figure 1 This is an overall flowchart of the main distribution network operation decision-making method based on health assessment provided in the embodiments of the present invention;

[0057] Figure 2 This is a flowchart illustrating the process of constructing a primary-secondary collaborative health benchmark profile and health scoring rules in an embodiment of the present invention;

[0058] Figure 3 This is a flowchart illustrating the process of obtaining the current health assessment result containing the collaborative health deviation distribution in an embodiment of the present invention;

[0059] Figure 4 This is a flowchart illustrating the measurement configuration optimization based on the collaborative health deviation distribution in this embodiment of the invention.

[0060] Figure 5 This is a flowchart of weighted state estimation and parameter identification based on health deviation weights in an embodiment of the present invention;

[0061] Figure 6 This is a pre-generated flowchart of the multi-agent collaborative decision-making model in this embodiment of the invention;

[0062] Figure 7 This is a flowchart illustrating the generation of a primary-secondary collaborative operation strategy in an embodiment of the present invention;

[0063] Figure 8 This is a flowchart of the incremental learning and adjustment process of the collaborative decision-making model in this embodiment of the invention;

[0064] Figure 9 This is a structural block diagram of the main distribution network operation decision-making system based on health assessment provided in an embodiment of the present invention. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0066] In existing power grid operation decision-making systems, both measurement configuration and state estimation weights adopt a static, fixed model, which cannot adapt to the dynamic fluctuations in power grid operation. This results in insufficient state estimation accuracy, thus affecting the accuracy and reliability of operation decisions. To address this issue, this embodiment provides a method and system for power grid operation decision-making based on health assessment. By constructing a health benchmark profile and scoring rules to quantify differences in power grid operating states, the method dynamically optimizes measurement configuration and adjusts state estimation weights accordingly. This tilts measurement resources towards key areas and strengthens the role of data from high-bias areas in state estimation. By improving state estimation accuracy from both data source and calculation mechanism perspectives, the method ultimately ensures the accuracy and reliability of power grid operation decision-making.

[0067] Example 1:

[0068] This invention proposes a method for decision-making in the coordinated operation of a distribution network based on health assessment, such as... Figure 1 As shown, the method includes:

[0069] Step S1: Based on the dual-end multi-source operation data of the main distribution coordinated power grid and the preset power grid health indicators, construct the health benchmark profile and health scoring rules of the main distribution coordinated power grid, and obtain the current health assessment results including the distribution of coordinated health deviation by combining real-time measurement data.

[0070] This step quantifies the differences in power grid operating states, providing an objective basis for subsequent measurement configuration optimization and state estimation weight adjustment. Specifically, it includes:

[0071] Step S11: Construct a health benchmark profile and health scoring rules for primary and secondary collaboration, such as... Figure 2 As shown, it includes the following steps:

[0072] Step S111: Based on the dual-end multi-source operation data of the main grid and distribution grid, extract the independent electrical quantity characteristics of the main grid side and the distribution grid side, as well as the coordinated characteristic quantities corresponding to the interaction status indicators of the main grid and distribution grid.

[0073] The system acquires dual-end, multi-source operational data from the SCADA measurement system, PMU measurement system, energy management system, and distribution automation system of the main grid and distribution network. This includes voltage amplitude and phase angle of each node on the main grid side, active and reactive power of each branch and electrical quantities of each side of the main transformer, as well as voltage, current, active and reactive power, output of distributed power sources, energy storage charging and discharging power, and interruptible load status of each feeder node on the distribution network side.

[0074] The collaborative feature quantities are calculated by combining the master-slave interaction status indicators. The calculation methods for the master-slave interaction status indicators and corresponding collaborative feature quantities are as follows:

[0075] Active power balance at the main grid and distribution grid: A virtual measurement point is set at the main grid and distribution grid. The algebraic sum of the active power of all tie lines at the measurement point is calculated (the flow from the main grid side to the distribution grid side is the positive direction) to obtain the active power exchange. The ratio of its absolute value to the total power generation of the main grid side or the total load of the distribution grid side is taken as the quantitative value.

[0076] Reactive power balance at the main grid and distribution grid: The reactive power exchange is obtained by calculating the algebraic sum of the reactive power of all tie lines at the virtual measurement point, and the ratio of its absolute value to the total power generation of the main grid or the total load of the distribution grid is taken as the quantitative value.

[0077] Consistency of bus voltage at checkpoints: The ratio of the standard deviation to the average value of the voltage amplitude of all checkpoints is taken. The smaller the ratio, the higher the consistency.

[0078] Voltage support strength of the main grid to the distribution network: Obtain the voltage change of the main grid side and the distribution network side bus at the virtual measurement point, and use the recursive least squares method to fit the transmission coefficient. The closer the coefficient is to 1, the stronger the support. Alternatively, calculate the ratio of the equivalent impedance of the main grid side to the equivalent load impedance of the distribution network side, and take its reciprocal as the support strength.

[0079] Reverse power transmission margin from distribution network to main network: Calculate the difference between the total output of distributed generation and the total load of distribution network, divide by the rated capacity of distribution network to obtain the reverse power transmission ratio, and a value greater than zero indicates that reverse power transmission is in progress;

[0080] Matching degree of main grid operation mode: The similarity between the current main grid operation mode and the load curve shape of the distribution network is calculated using cosine similarity or Pearson correlation coefficient.

[0081] Phase difference at the junction: Calculate the voltage phase angle of the equivalent busbar on the main grid side and the equivalent busbar on the distribution grid side respectively, and take the difference between the two.

[0082] Step S112: Based on historical operating data and the aforementioned collaborative characteristic quantities, determine the normal operating fluctuation range of each parameter on the main grid side, distribution grid side, and main-distribution interaction section, as well as the correlation between parameters, to form a health benchmark profile of main-distribution collaboration.

[0083] The historical operating data from the dual-end multi-source operating data is divided into several subsets according to season and day type. Each subset contains both routine electrical quantities and corresponding time-specific collaborative characteristic quantities. The quantile regression method is used to calculate the operating fluctuation range of each parameter under normal operating conditions. For example, the voltage amplitude is taken as the decimal and ninetieth percentile as the lower and upper boundaries of normal fluctuation.

[0084] The correlations between different parameters are calculated using canonical correlation analysis or covariance matrix calculations, including correlation coefficients between conventional electrical quantities, between conventional electrical quantities and synergistic characteristic quantities, and between synergistic characteristic quantities themselves. All fluctuation ranges and correlations are integrated to form a health benchmark profile of the primary and secondary components' coordination.

[0085] Step S113: Using the coordinated characteristic quantity as an adjustment factor, the preset power grid health indicators are coupled and weighted to determine the deviation level threshold, dynamic weight coefficient and weighted deduction calculation rules of each indicator, and form the health scoring rules for main and distribution coordination.

[0086] The preset power grid health indicators consist of conventional electrical quantity indicators (node ​​voltage deviation, branch power fluctuation, line and transformer load rate, etc.) and collaborative characteristic quantity indicators. Each indicator corresponds to a basic weight determined by expert experience or the analytic hierarchy process.

[0087] The basic weights are dynamically adjusted using collaborative characteristic quantities as adjustment factors: For collaborative characteristic quantity indicators, the weights are directly adjusted based on their real-time deviation, with higher weights for larger deviations; for conventional electrical quantity indicators, a correlation matrix between collaborative characteristic quantities and each indicator is pre-established. When the deviation of a certain collaborative characteristic quantity exceeds the normal range, the weight of the corresponding related indicator is increased according to the correlation strength, with a larger increase for higher correlation strength.

[0088] The degree to which real-time operating values ​​deviate from the normal fluctuation range in the health benchmark profile is divided into four levels: normal, slight deviation, moderate deviation, and severe deviation. Each level corresponds to a preset deduction value, with higher deviation levels resulting in higher deduction values. A weighted deduction model with a fixed full score is used to calculate the health score: the deduction value for each indicator's deviation level is multiplied by its dynamic weight coefficient to obtain an individual weighted deduction value. All individual deduction values ​​are summed and then deducted from the full score to obtain the final health score. A higher score indicates a better power grid health status.

[0089] Based on the collaborative health deviation distribution in the current health assessment results, high deviation spatial locations are screened and areas to be optimized are determined. The optimal measurement scheme is solved by combining candidate measurement points, and the measurement configuration optimization is completed to obtain the optimized measurement data.

[0090] Extract the collaborative health deviation distribution from the current health assessment results, set health deviation weights based on the spatial location of each measurement point after measurement configuration optimization, perform weighted state estimation based on the optimized measurement data and health deviation weights, and simultaneously use the state estimation residual sequence to identify parameters, thereby obtaining the state estimation results of the main and distribution coordinated power grid and the corrected network parameters.

[0091] Based on the state estimation results and the corrected network parameters, a short-term operating scenario is generated by combining the pre-generated multi-agent collaborative decision-making model, and collaborative decision-making is carried out to obtain a collaborative operating strategy that includes control instructions for each controllable device in the main and distribution networks.

[0092] Step S12: Combine real-time measurement data to obtain the current health assessment result, including the distribution of co-health deviation, such as... Figure 3 As shown, it includes the following steps:

[0093] Step S121: Extract real-time operating values ​​from the main network side, distribution network side, and main-distribution interaction section based on real-time measurement data. The data types of real-time measurement data and dual-end multi-source operating data are consistent.

[0094] Step S122: Compare the real-time operating value with the normal operating fluctuation range of the corresponding parameter in the health benchmark profile, calculate the deviation of each power grid health indicator, and correct the deviation by combining the deviation of the correlation between parameters.

[0095] The deviation rate is used to quantify the degree of deviation between the real-time operating value and the corresponding fluctuation range. The formula is as follows:

[0096] ;

[0097] Wherein, the half-width of the fluctuation range = (upper boundary of the normal operation fluctuation range - lower boundary of the normal operation fluctuation range) / 2.

[0098] A basic deviation is generated based on a linear mapping of the deviation rate; the larger the absolute value of the deviation rate, the larger the basic deviation. Simultaneously, the correlation between parameters in the health baseline profile is retrieved, and the absolute deviation of the real-time correlation coefficient relative to the historical baseline correlation coefficient is calculated as the correlation deviation degree. If the correlation deviation degree exceeds a preset threshold, a risk of mismatch is identified, and the basic deviation is corrected by matching a corresponding amplification factor based on the correlation deviation degree; the larger the correlation deviation degree, the larger the amplification factor.

[0099] Step S123: Based on the health scoring rules, perform weighted calculation and level mapping on the corrected deviation amount to obtain the current primary and secondary collaborative health score.

[0100] Step S124: Classify regions according to the main grid side, distribution grid side, and main-distribution interaction section. Statistically count the number of power grid health indicators, the average deviation, and the coordinated mismatch markers of different deviation levels in each region. Generate the coordinated health deviation distribution and combine it with the current main-distribution coordinated health score to form the current health assessment result.

[0101] Step S2: Based on the collaborative health deviation distribution in the current health assessment results, screen high deviation spatial locations and determine the areas to be optimized. Combine candidate measurement points to solve for the optimal measurement scheme, complete the measurement configuration optimization, and obtain the optimized measurement data.

[0102] This step dynamically optimizes measurement configuration in high-bias areas to fill measurement coverage gaps, such as... Figure 4 As shown, it includes the following steps:

[0103] Step S21: Select spatial locations from the collaborative health deviation distribution where the deviation exceeds the preset threshold (the threshold corresponds to a slight deviation), and mark their topological coordinates (the number of the node exceeding the standard on the main network side, the location of the feeder section and node exceeding the standard on the distribution network side, and the corresponding point of the point exceeding the standard on the main-distribution interaction section) and the corresponding deviation level, the type of the exceeding indicator, and the collaborative mismatch association information.

[0104] Step S22: Use the selected spatial locations and their electrical association ranges as the areas to be optimized: the association range of the nodes exceeding the standard on the main grid side includes adjacent branches and upstream and downstream associated nodes; the association range of the feeder sections exceeding the standard on the distribution network side includes the upstream and downstream extension sections of the feeder and associated branch feeders; the association range of the corresponding exceeding gate on the main-distribution interaction section includes the adjacent main grid nodes, distribution network nodes and corresponding tie lines on both sides of the gate.

[0105] Step S23: Set corresponding deviation weights based on the deviation of each spatial location within the area to be optimized. The higher the deviation and the higher the deviation level, the greater the deviation weight.

[0106] Step S24: With the goal of minimizing the weighted state estimation error of the region to be optimized, and with the location of the candidate measurement points and the collectable parameters as constraints, solve for the optimal measurement scheme that includes the installation location, type and quantity of measurement equipment.

[0107] Construct the optimization objective function:

[0108] ;

[0109] in, To optimize the objective function value, The total number of regions to be optimized. For the first The deviation weights of each region to be optimized For the first The state estimation error of a region to be optimized is the average or maximum value of the estimated variances of all state variables in that region.

[0110] The decision variables are the selection states of each candidate measurement point. Each candidate point corresponds to multiple decision variables, representing the installation selection of different types of measurement equipment. A genetic algorithm is used to solve the optimization problem: the decision variables are represented by binary codes, the fitness function is the negative of the optimization objective function, and the algorithm iteratively evolves through selection, crossover, and mutation operations until the maximum number of iterations is reached or the optimization magnitude of multiple consecutive generations is less than a preset threshold. After convergence, the optimal measurement scheme is output.

[0111] Step S25: Based on the optimal measurement scheme, generate supplementary pseudo-measurement data in the high-bias region, remove redundant measurement data in the low-bias region, and integrate to obtain the optimized measurement data.

[0112] For missing measurement points in high-deviation areas, pseudo measurement data is generated based on real-time data from nearby measurement points and load / distributed power generation output prediction data. For areas outside the area to be optimized, if the density of the same type of measurement points exceeds the preset threshold or the repeatability of parameters collected from adjacent points exceeds the threshold, redundant measurement data is removed.

[0113] Step S3: Extract the collaborative health deviation distribution from the current health assessment results, set health deviation weights based on the spatial location of each measurement point after measurement configuration optimization, perform weighted state estimation based on the optimized measurement data and health deviation weights, and simultaneously use the state estimation residual sequence to identify parameters, thereby obtaining the state estimation results and corrected network parameters of the main and distribution coordinated power grid.

[0114] This step improves the accuracy of state estimation from the perspective of computational mechanisms, while correcting errors in the power grid model parameters, such as... Figure 5 As shown, it includes the following steps:

[0115] Step S31: Based on the collaborative health deviation distribution corresponding to the spatial location of each measurement point, set the health deviation weight for each measurement point using gradient settings, and construct a diagonal matrix of health deviation weights. Measurement points with larger location deviations, higher deviation levels, or those exhibiting collaborative mismatch correlations will have higher health deviation weights.

[0116] Step S32: Input the optimized measurement data and the health deviation weight diagonal matrix into the weighted least squares state estimation algorithm, and iteratively solve to obtain the state estimation results of voltage amplitude, phase angle and branch power flow of all nodes in the main distribution coordinated power grid.

[0117] The iterative solution process is as follows:

[0118] 1. Based on the main distribution coordinated power grid topology and branch parameters, initialize the voltage amplitude and phase angle of each node as the initial values ​​for state variable iteration;

[0119] 2. Construct a diagonal matrix of health deviation weights;

[0120] 3. Based on the power flow mechanism of the power grid, establish the nonlinear measurement function equation, perform a first-order Taylor expansion on the current iterative state point, and solve the Jacobian matrix of the measurement equation with respect to the state variables;

[0121] 4. Integrate the Jacobian matrix and the health deviation weight matrix to form a weighted least squares information matrix and an equivalent residual constant vector, and simultaneously construct a normal solution system of equations;

[0122] 5. Solve the normal equations to obtain the state variable corrections, and update the state variables;

[0123] 6. Calculate the maximum change in state variables between the two iterations. If it is less than the preset convergence threshold, stop the iteration; otherwise, return to step 3 to continue the iteration.

[0124] Step S33: Extract the residual time series of each measurement point during the state estimation process (the difference between the actual measured value and the theoretical calculated value of the state estimation), and combine it with historical measurement data to identify the line impedance, transformer equivalent parameters and main distribution interaction coupling parameters through the Kalman filter algorithm to obtain the corrected network parameters and replace the original offline static tuning parameters.

[0125] Step S4: Based on the state estimation results and the corrected network parameters, generate a short-term operating scenario and make collaborative decisions in conjunction with the pre-generated multi-agent collaborative decision-making model to obtain a collaborative operation strategy that includes control instructions for each controllable device in the main distribution network.

[0126] This step, based on accurate state awareness results, enables primary and secondary collaborative optimization decisions, specifically including:

[0127] Step S41: Pre-generate a multi-agent cooperative decision-making model, such as Figure 6As shown, it includes the following steps:

[0128] Step S411: Generate a sample set of running scenarios with health labels:

[0129] 1. Based on the load level, distributed power output level and network topology in the historical operation data, and combined with the corresponding health assessment results, a clustering algorithm is used to divide the historical operation status into several typical operation categories;

[0130] 2. Extract the representative operational cross-sections of each category that are closest to the cluster center, add corresponding health labels, and generate a basic operational mode sample set;

[0131] 3. Based on preset extreme event data (fault timing, meteorological change curves, load change curves, etc.), modify the corresponding parameters of the base sample to generate continuous time-series dynamic scenario samples from normal to extreme and then back to normal.

[0132] 4. Integrate the basic operation mode samples and dynamic scenario samples to obtain the operation scenario sample set.

[0133] Step S412: Based on the operational scenario sample set, extract the operational status characteristics and health assessment results of the main grid and distribution network coordinated network, and map them to the state spaces of the main grid intelligent agent and each distribution network feeder intelligent agent, respectively.

[0134] The operational status characteristics include main grid-side node voltage deviation, main transformer load rate, and switching power deviation at the gateway; distribution network-side feeder node voltage deviation, line load rate, three-phase imbalance, distributed power output deviation, energy storage state of charge and charging / discharging power; as well as main grid-distribution coordinated health score and coordinated health deviation distribution. After normalization of each characteristic, they are concatenated into a state vector. The main grid agent state space includes main grid-side and gateway-related features, while the distribution network feeder agent state space includes the features of the corresponding feeder and its neighboring nodes. All agents include global health scores and coordinated deviation distribution characteristics.

[0135] Step S413: Define the discrete and continuous action spaces of each intelligent agent based on the type and physical regulation range of controllable equipment in the main distribution coordinated power grid:

[0136] Main grid intelligent agent action space: up / down / holding of tap positions of on-load tap changer, and switching commands of main grid side capacitor banks;

[0137] The action space of the distribution network feeder intelligent agent includes: the number of switching groups of the reactive power compensation device at the feeder outlet, the charging and discharging power command of the energy storage system, and the disconnection ratio of interruptible loads.

[0138] Step S414: Construct a multi-objective reward function that includes a line operation safety penalty, a network loss negative reward, and a health score change reward:

[0139] Line operation safety penalty items: Penalties for exceeding the safety threshold for line load rate and exceeding the limit for node voltage;

[0140] Negative incentive for network loss: A negative incentive is given for active network loss;

[0141] Health score change reward: (current health score - previous health score) × preset positive coefficient. A positive reward is given for an increase in health score, and a negative reward is given for a decrease in health score.

[0142] The weight coefficients of each sub-item are configured according to the proportion of each type of scenario in the sample set of operating scenarios to ensure the balance of the magnitude of the reward function.

[0143] Step S415: Use a multi-agent reinforcement learning algorithm to jointly train the main network agent and each distribution network feeder agent until the multi-objective reward function converges, and obtain the collaborative decision-making model.

[0144] Each agent is configured with a policy network and a value network containing recurrent neural network layers. The training process is as follows:

[0145] 1. Randomly select initial time-series samples from the running scenario sample set as the initial state of the simulation environment;

[0146] 2. Each agent outputs an action command based on its current state. The simulation environment executes the joint action, updates its state, and calculates the reward value.

[0147] 3. Store the time sequence, action, reward, and next time sequence into the shared experience pool;

[0148] 4. Randomly sample experience data from the experience pool to update the value network and policy network of each agent;

[0149] 5. Continue iterating until the fluctuation range of the reward value is less than the preset convergence threshold for multiple consecutive rounds, and save the optimal policy network parameters.

[0150] Step S42: Generate a cooperative operation strategy, such as Figure 7 As shown, it includes the following steps:

[0151] Step S421: Using the current state estimation result as the initial operating section, and combining the load forecast data and distributed power output forecast data for the future preset time period, perform time-by-time power flow simulation using the corrected network parameters to generate a short-term operating scenario sequence containing multiple time sections.

[0152] Step S422: Input the short-term operation scenario sequence into the pre-generated collaborative decision-making model. The main grid agent outputs main grid-side voltage regulation and reactive power compensation control commands, and each distribution network feeder agent outputs distribution network-side reactive power regulation, energy storage charging and discharging, and interruptible load control commands. Integrate the control commands of all agents to form a main grid-distribution collaborative operation strategy that meets the timing constraints.

[0153] Step S43: Incremental learning and adjustment of the collaborative decision-making model, such as... Figure 8 As shown, it includes the following steps:

[0154] After obtaining and executing the collaborative operation strategy, perform the following steps:

[0155] S431: Collect actual measurement data within the first sampling period after the execution of the collaborative operation strategy, calculate the actual health score after execution according to the health scoring rules, and compare it with the health score before execution to obtain the change in health score.

[0156] S432: Construct incremental samples based on the running status before execution, the control instructions in the collaborative operation strategy, and the changes in health scores after execution.

[0157] S433: Employing a mini-batch gradient descent algorithm, combined with incremental samples and randomly drawn samples from the experience pool, the policy network and value network of the collaborative decision-making model are updated in a finite number of steps with the goal of minimizing the prediction error of changes in health scores.

[0158] Example 2:

[0159] like Figure 9 As shown, this invention provides a primary-distribution coordinated distribution network operation decision-making system based on health assessment. This system is used to implement the primary-distribution coordinated distribution network operation decision-making method based on health assessment described in Embodiment 1 above, specifically including:

[0160] The health assessment module is used to construct a health benchmark profile and health scoring rules for the main and distribution coordinated power grid based on the dual-end multi-source operation data of the main and distribution coordinated power grid and the preset power grid health indicators, and to obtain the current health assessment results including the distribution of coordinated health deviations by combining real-time measurement data.

[0161] The measurement optimization module is used to screen high-deviation spatial locations and determine the area to be optimized based on the collaborative health deviation distribution in the current health assessment results, solve the optimal measurement scheme by combining candidate measurement points, complete the measurement configuration optimization, and obtain the optimized measurement data.

[0162] The identification module is used to extract the collaborative health deviation distribution in the current health assessment results, set the health deviation weights based on the spatial location of each measurement point after the measurement configuration optimization, perform weighted state estimation based on the optimized measurement data and health deviation weights, and simultaneously use the state estimation residual sequence to identify parameters, thereby obtaining the state estimation results of the main and distribution coordinated power grid and the corrected network parameters.

[0163] The collaborative decision-making module, based on the state estimation results and the corrected network parameters, combines the pre-generated multi-agent collaborative decision-making model to generate short-term operating scenarios and make collaborative decisions, thereby obtaining a collaborative operating strategy that includes control instructions for each controllable device in the main and distribution networks.

[0164] This embodiment provides a primary-distribution coordinated distribution network operation decision-making system based on health assessment, used to implement the aforementioned primary-distribution coordinated distribution network operation decision-making method based on health assessment. Therefore, the specific implementation of the primary-distribution coordinated distribution network operation decision-making system based on health assessment can be found in the previous embodiment section of the primary-distribution coordinated distribution network operation decision-making method based on health assessment. For example, the health assessment module, measurement optimization module, operation identification module, and coordinated decision-making module are used to implement steps S1, S2, S3, and S4 in the aforementioned primary-distribution coordinated distribution network operation decision-making method based on health assessment, respectively. Therefore, its specific implementation can be referred to the description of the corresponding embodiments. To avoid redundancy, it will not be repeated here.

[0165] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0166] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations 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, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0167] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0168] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for decision-making in the coordinated operation of a distribution network based on health assessment, characterized in that, include: Based on the dual-end multi-source operation data of the main distribution coordinated power grid and the preset power grid health indicators, a health benchmark profile and health scoring rules for the main distribution coordinated power grid are constructed, and the current health assessment results including the distribution of coordinated health deviation are obtained by combining real-time measurement data. Based on the collaborative health deviation distribution in the current health assessment results, high deviation spatial locations are screened and areas to be optimized are determined. The optimal measurement scheme is solved by combining candidate measurement points, and the measurement configuration optimization is completed to obtain the optimized measurement data. Extract the collaborative health deviation distribution from the current health assessment results, set health deviation weights based on the spatial location of each measurement point after measurement configuration optimization, perform weighted state estimation based on the optimized measurement data and health deviation weights, and simultaneously use the state estimation residual sequence to identify parameters, thereby obtaining the state estimation results of the main and distribution coordinated power grid and the corrected network parameters. Based on the state estimation results and the corrected network parameters, a short-term operating scenario is generated by combining the pre-generated multi-agent collaborative decision-making model, and collaborative decision-making is carried out to obtain a collaborative operating strategy that includes control instructions for each controllable device in the main and distribution networks.

2. The main distribution network operation decision-making method based on health assessment according to claim 1, characterized in that, The method constructs a health benchmark profile and health scoring rules for main-distribution coordination based on dual-end multi-source operation data of the main-distribution coordinated power grid and preset power grid health indicators, including: Based on the dual-end multi-source operation data of the main grid and distribution grid, we extract the independent electrical quantity characteristics of the main grid side and the distribution grid side, as well as the coordinated characteristic quantities corresponding to the interaction status indicators of the main grid and distribution grid. Based on historical operating data and the aforementioned collaborative characteristic quantities, the normal operating fluctuation range of each parameter on the main grid side, distribution grid side, and main-distribution interaction section and the correlation between parameters are determined, forming a health benchmark profile of main-distribution collaboration. The pre-set power grid health indicators are coupled and weighted using the aforementioned collaborative characteristic quantities as adjustment factors to determine the deviation level threshold, dynamic weight coefficient, and weighted deduction calculation rules for each indicator, thereby forming a health scoring rule for main and distribution coordination.

3. The main distribution network operation decision-making method based on health assessment according to claim 1, characterized in that, The process of obtaining the current health assessment result, which includes the distribution of collaborative health deviations, by combining real-time measurement data includes: Real-time operating values ​​are extracted from the main grid side, distribution network side, and main-distribution interaction section based on real-time measurement data; By comparing the real-time operating values ​​with the normal operating fluctuation range of the corresponding parameters in the health benchmark profile, the deviation of each power grid health indicator is calculated, and the deviation is corrected by combining the deviation of the correlation between parameters. Based on the aforementioned health scoring rules, the corrected deviation is weighted and mapped to a level to obtain the current primary and secondary collaborative health score. The deviation distribution of each power grid health indicator is statistically analyzed according to spatial region to generate a coordinated health deviation distribution, which is then combined with the current main and distribution coordinated health score to form the current health assessment result.

4. The main distribution network operation decision-making method based on health assessment according to claim 1, characterized in that, Based on the collaborative health deviation distribution in the current health assessment results, high-deviation spatial locations are screened and areas to be optimized are determined. The optimal measurement scheme is then calculated by combining candidate measurement points to complete the measurement configuration optimization and obtain optimized measurement data, including: Spatial locations whose deviation exceeds a preset threshold are selected from the collaborative health deviation distribution, and the spatial locations and their electrical associated ranges are used as the areas to be optimized. Based on the deviation of each spatial location within the area to be optimized, corresponding deviation weights are set. With the goal of minimizing the weighted state estimation error of the area to be optimized, and with the location and collectable parameters of the candidate measurement points as constraints, the optimal measurement scheme, which includes the installation location, type and quantity of measurement equipment, is obtained. Based on the optimal measurement scheme, supplementary pseudo-measurement data is generated in the high-bias region, redundant measurement data is removed in the low-bias region, and the optimized measurement data is obtained by integration.

5. The main distribution network operation decision-making method based on health assessment according to claim 1, characterized in that, The process involves extracting the coordinated health deviation distribution from the current health assessment results, setting health deviation weights based on the spatial location of each measurement point after measurement configuration optimization, performing weighted state estimation based on the optimized measurement data and health deviation weights, and simultaneously using the state estimation residual sequence for parameter identification to obtain the state estimation results and corrected network parameters of the main and distribution coordinated power grid, including: Based on the collaborative health deviation distribution corresponding to the spatial location of each measurement point, the health deviation weight of each measurement point is set by gradient, and a diagonal matrix of health deviation weights is constructed. The optimized measurement data and the health deviation weighted diagonal matrix are input into the weighted least squares state estimation algorithm, and the state estimation results of voltage amplitude, phase angle and branch power flow of all nodes in the main distribution coordinated power grid are obtained by iterative solution. The residual time series of each measurement point during the state estimation process is extracted. Combined with historical measurement data, the Kalman filter algorithm is used to identify the line impedance, transformer equivalent parameters, and main distribution interface coupling parameters to obtain the corrected network parameters.

6. The main distribution network operation decision-making method based on health assessment according to claim 1, characterized in that, The pre-generation process of the multi-agent cooperative decision-making model is as follows: Based on a pre-generated sample set of operating scenarios with health tags, the operating status characteristics and health assessment results of the main grid and distribution network are extracted and mapped to the state spaces of the main grid agent and each distribution network feeder agent, respectively. Based on the type and physical regulation range of controllable equipment in the main distribution coordinated power grid, the discrete and continuous action spaces of each intelligent agent are defined. Construct a multi-objective reward function that includes a line operation safety penalty item, a network loss negative reward item, and a health score change reward item; A multi-agent reinforcement learning algorithm is used to jointly train the main network agent and the agents of each distribution network feeder until the multi-objective reward function converges, thus obtaining a collaborative decision-making model.

7. The main distribution network operation decision-making method based on health assessment according to claim 6, characterized in that, The process of generating the sample set of the operating scenarios is as follows: Based on the load level, distributed power output level and network topology in historical operation data, and combined with the corresponding health assessment results, a clustering algorithm is used to divide the network into several typical operation categories. Extract representative operational cross-sections of each category and add health labels to generate a sample set of basic operational modes; Based on extreme event data, the basic samples are augmented with time series data to generate dynamic scene samples that include scenarios such as faults and sudden weather changes. By integrating basic operation mode samples and dynamic scenario samples, an operation scenario sample set is obtained.

8. The main distribution network operation decision-making method based on health assessment according to claim 1, characterized in that, Based on the state estimation results and the corrected network parameters, a short-term operating scenario is generated and collaborative decision-making is performed using a pre-generated multi-agent collaborative decision-making model. This process yields a collaborative operating strategy that includes control commands for each controllable device in the main and distribution networks, including: Using the current state estimation result as the initial operating section, combined with the load forecast data and distributed power output forecast data for the future preset time period, the corrected network parameters are used to perform time-by-time power flow simulation to generate a short-term operating scenario sequence. The short-term operation scenario sequence is input into the pre-generated collaborative decision-making model. The main grid agent outputs the main grid side voltage regulation and reactive power compensation control commands, and each distribution network feeder agent outputs the distribution network side reactive power regulation, energy storage charging and discharging and interruptible load control commands. Integrate the control commands of all intelligent agents to form a master-slave collaborative operation strategy that satisfies timing constraints.

9. The main distribution network operation decision-making method based on health assessment according to claim 1, characterized in that, After obtaining the collaborative operation strategy, it also includes: Collect actual measurement data after the collaborative operation strategy is executed, calculate the actual health score after execution according to the health scoring rules, and compare it with the health score before execution to obtain the change in health score. Incremental samples are constructed based on the pre-execution running status, control instructions in the collaborative operation strategy, and changes in health scores after execution. We employ a mini-batch gradient descent algorithm, combined with incremental samples, to incrementally learn and update the policy network and value network of the collaborative decision-making model.

10. A primary-distribution coordinated distribution network operation decision-making system based on health assessment, characterized in that, The system is used to implement the main distribution network operation decision-making method based on health assessment as described in any one of claims 1 to 9, specifically including: The health assessment module is used to construct a health benchmark profile and health scoring rules for the main and distribution coordinated power grid based on the dual-end multi-source operation data of the main and distribution coordinated power grid and the preset power grid health indicators, and to obtain the current health assessment results including the distribution of coordinated health deviations by combining real-time measurement data. The measurement optimization module is used to screen high-deviation spatial locations and determine the area to be optimized based on the collaborative health deviation distribution in the current health assessment results, solve the optimal measurement scheme by combining candidate measurement points, complete the measurement configuration optimization, and obtain the optimized measurement data. The identification module is used to extract the collaborative health deviation distribution in the current health assessment results, set the health deviation weights based on the spatial location of each measurement point after the measurement configuration optimization, perform weighted state estimation based on the optimized measurement data and health deviation weights, and simultaneously use the state estimation residual sequence to identify parameters, thereby obtaining the state estimation results of the main and distribution coordinated power grid and the corrected network parameters. The collaborative decision-making module, based on the state estimation results and the corrected network parameters, combines the pre-generated multi-agent collaborative decision-making model to generate short-term operating scenarios and make collaborative decisions, thereby obtaining a collaborative operating strategy that includes control instructions for each controllable device in the main and distribution networks.

Images