A method and system for optimizing energy saving and loss reduction of distributed photovoltaic active distribution network

Abstract

An optimization method for energy saving and loss reduction in an active distribution network with distributed photovoltaic (PV) power includes the following steps: Unified scheduling of distributed PV power output through an active coordination strategy, generating PV power output prediction values ​​and correcting their errors to form a PV power output curve; Based on the PV power output curve, analyzing the real-time energy flow status in the distribution network, and dynamically adjusting the charging and discharging strategies of PV power output and energy storage in conjunction with the load demand and voltage level of the distribution network nodes to optimize the energy transmission path; Based on the analysis results of the PV power output curve and energy storage strategy, constructing a multi-objective optimization model to comprehensively weigh the current state and future operating needs of the distribution network, and dynamically adjusting the weights between various optimization objectives; Applying the results of the multi-objective optimization to actual distribution network operation, correcting the PV power output curve and optimization strategy through real-time monitoring and feedback, forming a self-learning closed loop, and optimizing the active distribution network with distributed PV power.

Description

Technical Field

[0001] The present invention belongs to the technical field of power grid energy conservation, and more specifically, relates to an energy conservation and loss reduction optimization method and system for a distributed photovoltaic active distribution network. Background Art

[0002] With the rapid development and widespread application of distributed photovoltaic technology, the access volume of distributed photovoltaic in active distribution networks continues to increase. This trend provides new opportunities for the use of renewable energy. However, it also brings multiple challenges to the energy saving and loss reduction of distribution networks. The energy saving and loss reduction optimization methods in existing technologies usually focus on the scheduling and management of traditional power sources, and fail to fully consider the characteristics of distributed photovoltaics, such as power generation volatility and intermittency. This leads to poor optimization effects in practical applications, and cannot effectively improve the self-use rate of photovoltaic power generation or reduce the charging and discharging losses of energy storage systems.

[0003] In addition, the shortcomings of existing technologies in real-time monitoring and feedback make it impossible to respond to changes in distribution network load demand and PV output in a timely manner, further affecting the overall operating efficiency. Existing methods often rely on static models, lack flexibility, and cannot adapt to dynamically changing power demand and supply conditions. These problems highlight the urgent need for more comprehensive and dynamic optimization methods to achieve efficient utilization and management of distributed PV in active distribution networks.

[0004] Prior art document 1 (CN117833320A) discloses an optimization scheduling method and system for energy storage in a distributed photovoltaic distribution network, which belongs to the field of power system optimization technology; the pre-obtained predicted load power data and the predicted power generation data of the photovoltaic system are used as the initialization parameters of the pre-obtained energy storage optimization scheduling model, and the genetic algorithm is used to solve the energy storage optimization scheduling model to obtain an energy storage optimization scheduling plan with the minimum grid loss of the substation system and the minimum power fluctuation of the grid-connected nodes of the substation system. By comprehensively considering the load power and photovoltaic power generation power, an energy storage optimization scheduling model is established, and the energy storage optimization scheduling model is solved by a genetic algorithm with strong global optimization ability to obtain an energy storage optimization scheduling plan with the minimum grid loss of the substation system and the minimum power fluctuation of the grid-connected nodes of the substation system.

[0005] The shortcomings of the prior art document 1 are that the optimization scheduling model mainly relies on pre-acquired load power and photovoltaic power generation data, resulting in insufficient response of the model to dynamic changes in actual operation; it mainly focuses on minimizing network losses and power fluctuations, and lacks comprehensive consideration of the photovoltaic power generation self-use rate and the charging and discharging losses of the energy storage system; it fails to make full use of historical data and real-time feedback for self-adjustment, which may lead to attenuation of the effect in long-term operation. Summary of the invention

[0006] In order to solve the deficiencies in the prior art, the present invention provides an energy-saving and loss-reducing optimization method and system for a distributed photovoltaic active distribution network, aiming to improve the energy efficiency of the distribution network and reduce losses.

[0007] The present invention adopts the following technical solution.

[0008] A first aspect of the present invention provides a method for optimizing energy saving and loss reduction of a distributed photovoltaic active distribution network, comprising the following steps:

[0009] Through active coordination strategy, the distributed photovoltaic output is uniformly dispatched to generate photovoltaic output forecast value, and the error of photovoltaic output forecast value is corrected to form photovoltaic output curve;

[0010] According to the photovoltaic output curve and the monitoring value of the distribution network, the real-time flow state of energy in the distribution network is extracted to correct the photovoltaic output curve and adjust the charging and discharging strategy of the energy storage equipment;

[0011] According to the revised photovoltaic output curve and the adjusted charging and discharging strategy of the energy storage equipment, a multi-objective optimization model is constructed with the goals of maximizing photovoltaic output, minimizing energy storage charging and discharging efficiency, and minimizing operating costs. When training the multi-objective optimization model, the weights corresponding to each objective are dynamically adjusted according to the actual operating parameters of the distribution network, the load demand forecast value, and the energy supply change forecast value.

[0012] Based on the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment generated by the trained multi-objective optimization model, the deviation between the actual photovoltaic output and the photovoltaic output scheduling plan is monitored in real time. According to the changes in the node load and voltage of the distribution network, the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment are updated to control the operation of the distribution network.

[0013] Preferably, the distributed photovoltaic output is uniformly dispatched through an active coordination strategy to generate a photovoltaic output prediction value, and the photovoltaic output prediction value error is corrected to form a photovoltaic output curve, including:

[0014] Collect actual output data of distributed photovoltaics, operating status data of distribution networks, and user load demand data, and generate photovoltaic output forecast values ​​based on historical meteorological data;

[0015] Obtain historical photovoltaic output data; use statistical methods to fit the errors of historical photovoltaic output data, and correct the photovoltaic output prediction value based on the fitted errors.

[0016] Preferably, historical photovoltaic output data is obtained; errors of the historical photovoltaic output data are fitted using a statistical method, and the photovoltaic output prediction value is corrected based on the fitted errors, including:

[0017] The probability distribution of the error of the historical photovoltaic output data in different power ranges is fitted based on the Gaussian mixture model, and the error of the photovoltaic output prediction value is obtained from the probability distribution of the error using the Monte Carlo sampling method.

[0018] According to the time sequence, the sum of the photovoltaic output prediction value and the photovoltaic output prediction value is used as the corrected photovoltaic output to form a photovoltaic output curve.

[0019] Preferably, according to the photovoltaic output curve, combined with the monitoring value of the distribution network, the real-time flow state of energy in the distribution network is extracted to correct the photovoltaic output curve and adjust the charging and discharging strategy of the energy storage device, including:

[0020] Extract characteristic values ​​and corresponding time of characteristic values ​​on the photovoltaic output curve, where the characteristic values ​​include peak value, valley value and change rate;

[0021] The distribution network monitoring values ​​at the time corresponding to the characteristic values ​​are used, where the distribution network monitoring values ​​include the energy storage power monitoring value, the node load demand monitoring value, and the grid loss monitoring value, to extract the real-time flow state of energy in the distribution network as follows:

[0022] ,

[0023] In the formula, Indicates that the eigenvalue corresponds to time Energy storage power monitoring value under Indicates that the eigenvalue corresponds to time The node load demand monitoring value under Indicates that the eigenvalue corresponds to time The power grid loss monitoring value under is the characteristic value;

[0024] According to the real-time load demand of each node, the voltage level of each node is monitored and adjusted when the voltage exceeds the set range. According to the real-time flow state of energy, the charging and discharging strategy of photovoltaic output and energy storage equipment is dynamically adjusted through the energy balance principle; the setting rules are as follows:

[0025] When the photovoltaic output is greater than the node load demand, the energy storage device is charged first;

[0026] When the photovoltaic output is less than the node load demand, the energy storage device discharges first to supplement the load.

[0027] Preferably, a multi-objective optimization model is constructed based on the corrected photovoltaic output curve and the adjusted charging and discharging strategy of the energy storage device, with the objectives of maximizing photovoltaic output, minimizing energy storage charging and discharging efficiency, and minimizing operating costs. When the multi-objective optimization model is trained, the weights corresponding to the various objectives are dynamically adjusted according to the actual operating parameters of the distribution network, the load demand forecast value, and the energy supply change forecast value, including:

[0028] The self-use rate of maximizing photovoltaic output is calculated as follows:

[0029] ,

[0030] Where:

[0031] Indicates the self-use rate of photovoltaic output; Indicates the power of photovoltaic output directly used for load; Indicates the total photovoltaic output power;

[0032] Minimize the energy storage charging and discharging loss by the following formula

[0033] ,

[0034] Where:

[0035] Indicates energy storage loss; Indicates time Charging power of energy storage; Indicates time Discharge power of energy storage; Indicates the energy storage charging efficiency; Indicates the energy storage discharge efficiency;

[0036] Minimize operating cost by calculating

[0037] ,

[0038] Where:

[0039] represents the total operating cost; Indicates time The power of purchased electricity; Indicates time the price of purchased electricity; Indicates time Equipment maintenance costs.

[0040] Preferably, the comprehensive objective function is calculated as follows:

[0041] ,

[0042] Where:

[0043] represents the comprehensive optimization objective; , and are the weight factors for each target respectively.

[0044] Preferably, dynamically adjusting the weights corresponding to various objectives includes:

[0045] In the multi-objective optimization model, each particle is regarded as a different weight allocation scheme; each particle represents a weight configuration, and the fitness is evaluated by the objective functions including maximizing photovoltaic output, minimizing energy storage efficiency, and minimizing operating costs;

[0046] Construct a fitness function, evaluate the fitness value of each particle through the specific parameters of photovoltaic output, energy storage efficiency and operating cost, and obtain the optimal weight combination based on the actual operating conditions to balance the three objectives;

[0047] Each particle adjusts the weight distribution through its own historical optimal position and global optimal position, and dynamically adjusts the weight configuration of the multi-objective optimization model; through the iterative optimization of the particle swarm algorithm, the best weight distribution scheme is gradually obtained;

[0048] After the particle swarm algorithm converges to the global optimal solution, the final weight configuration is determined, and the optimization process of photovoltaic output, energy storage charging and discharging strategy and operating cost is adjusted based on the final weight configuration to achieve dynamic adjustment of the weights corresponding to each target.

[0049] Preferably, based on the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage device generated by the trained multi-objective optimization model, the deviation between the actual photovoltaic output and the photovoltaic output scheduling plan is monitored in real time, and the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage device are updated according to the changes in the node load and voltage of the distribution network to control the operation of the distribution network, including:

[0050] Through the multi-objective optimization model, the optimal operation strategy of the distribution network is obtained, and the optimization results are applied to the actual operation of the distribution network, including: adjusting the power of photovoltaic output, optimizing the charging and discharging strategy of the energy storage system, and coordinating the purchase of electricity to achieve the balance between the load demand and power supply of each node in the distribution network;

[0051] During the actual operation of the distribution network, the actual power output of the photovoltaic output is monitored in real time through the photovoltaic output curve, and the operation deviation of the distribution network is judged in combination with the node load demand and the status of the energy storage equipment; the operation strategy is corrected in real time according to the changes in the photovoltaic output curve;

[0052] For the detected distribution network operation deviation, automatic adjustment is carried out through the built-in feedback mechanism. Based on historical data and real-time monitoring data, the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment are gradually corrected through a self-learning closed loop;

[0053] When all optimization targets of the distribution network in actual operation deviate from the predicted values ​​within the set range, the self-learning closed loop completes the optimization task; in future operations, when new changes and anomalies are detected, the next round of optimization and adjustment will be automatically triggered.

[0054] Preferably, the operation deviation of the distribution network is adjusted through a feedback mechanism, including:

[0055] When there is a difference between the photovoltaic output curve and the predicted value, the difference between the generated photovoltaic output prediction value and the actual corrected photovoltaic output curve is analyzed, and the charging and discharging behavior of the energy storage device is adjusted to match the current power demand with the photovoltaic output curve;

[0056] When the node load demand in the distribution network changes, based on the analysis of the photovoltaic output curve and the node load demand and voltage level of the distribution network, the energy dispatch of the energy storage device is dynamically optimized, and the charging and discharging timing is reasonably adjusted to maintain the energy flow balance of each node in the power grid;

[0057] According to the adjustment of future operating needs, the weight of each optimization target is re-evaluated through the multi-objective optimization model to adapt to the changes in distribution network operation. Through weight adjustment, the dispatch strategy of energy storage equipment is optimized under different operating conditions.

[0058] The second aspect of the present invention provides an energy-saving and loss-reduction optimization system for a distributed photovoltaic active distribution network, including: a photovoltaic output scheduling module, an energy flow analysis module, a multi-objective optimization model module, and a real-time monitoring and feedback module;

[0059] The photovoltaic output dispatching module is used to uniformly dispatch distributed photovoltaic output through active coordination strategies, generate photovoltaic output forecast values ​​using historical meteorological data and real-time monitoring information, and correct the forecast error to form a photovoltaic output curve;

[0060] The energy flow analysis module is used to monitor the operating parameters of the distribution network in real time. Combined with the corrected photovoltaic output curve, it extracts the real-time flow status of energy in the distribution network. Based on the monitoring results, it dynamically adjusts the photovoltaic output and the charging and discharging strategy of the energy storage equipment to maintain the balance of energy supply and demand.

[0061] The multi-objective optimization model module is used to build and train a multi-objective optimization model based on the objectives of maximizing photovoltaic output, minimizing energy storage charging and discharging efficiency, and minimizing operating costs. During the training process, the weights of each optimization objective are dynamically adjusted according to the actual operating parameters of the distribution network, load demand forecast values, and energy supply change forecast values.

[0062] The real-time monitoring and feedback module is used to monitor the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment in real time, monitor the deviation between the actual photovoltaic output and the scheduling plan, and timely update the photovoltaic output scheduling plan and the charging and discharging strategy of the energy storage equipment according to the changes in the node load and voltage of the distribution network.

[0063] Compared with the prior art, the beneficial effects of the present invention include at least: the present invention can realize energy-saving and loss-reduction optimization of the distribution network, improve the energy efficiency of the distribution network and reduce losses through the implementation of the distributed photovoltaic active distribution network energy-saving and loss-reduction optimization method and system. The present invention provides strong support for the sustainable development of the distribution network and has important practical application value.

[0064] In order to accurately track the fluctuation of distributed photovoltaic output, an optimization method for energy saving and loss reduction of active distribution network containing distributed photovoltaic is proposed. The difference in probability distribution of prediction error when photovoltaic power prediction value is in different ranges is analyzed through historical data, and the probability density distribution of photovoltaic power prediction error is fitted by Gaussian mixture model. A distributed photovoltaic prediction error model is established to accurately characterize the distributed photovoltaic prediction error, track the fluctuation of distributed photovoltaic output, and obtain the distributed photovoltaic power output scenario that fits the actual operating status, which is conducive to the accurate characterization of voltage and loss distribution and the formulation of energy saving and loss reduction strategies.

[0065] Taking into account the load type of the distribution system, a comprehensive load model for energy-saving and voltage-reduction control is established. Through the comprehensive load model for energy-saving and voltage-reduction control, the load consumption is expressed as a function of voltage, and the connection between load power and voltage is established. Under the premise of ensuring the quality of energy supply, the voltage of the distribution transformer feeder can be strategically reduced to reduce the overall energy consumption of the system, and energy saving and loss reduction can be achieved by voltage regulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is a flow chart of a method for optimizing energy saving and loss reduction of a distributed photovoltaic active distribution network provided in accordance with an embodiment of the present invention. DETAILED DESCRIPTION

[0067] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0068] like Figure 1 As shown, the embodiment 1 of the present invention provides an energy-saving and loss-reducing optimization method for a distributed photovoltaic active distribution network, comprising the following steps:

[0069] Step 1: Distributed photovoltaic output is uniformly dispatched through an active coordination strategy to generate a photovoltaic output forecast value, and the error of the photovoltaic output forecast value is corrected to form a photovoltaic output curve;

[0070] Preferably, step 1 comprises:

[0071] Collect distributed photovoltaic output data, distribution network operation status data and user demand data, process the data, obtain historical measured meteorological data, obtain photovoltaic output forecast values ​​based on the historical measured meteorological data, correct the photovoltaic forecast error, and obtain the photovoltaic output curve after error correction.

[0072] Preferably, the data processing includes:

[0073] Data cleaning, removing outliers and noise data;

[0074] Data transformation, converting raw data into a format suitable for processing by optimization algorithms;

[0075] Data association analysis is performed to establish a correlation model between distributed photovoltaic output, distribution network operating status and user demand.

[0076] Preferably, obtaining the photovoltaic output curve after error correction includes:

[0077] Obtain historical measured meteorological data, and obtain historical photovoltaic output forecast values ​​based on the historical measured meteorological data;

[0078] Obtain photovoltaic output prediction error based on historical photovoltaic output prediction values ​​and historical photovoltaic output data;

[0079] Based on the historical PV output data, the Gaussian mixture model is used to fit the probability distribution of PV output prediction errors in different power ranges, and the Monte Carlo simulation method is used to sample the PV output prediction errors from the obtained distribution.

[0080] The prediction error of photovoltaic output is calculated as follows:

[0081]

[0082] Where:

[0083] Indicates time Photovoltaic output curve under Indicates the initial output calculated based on meteorological data; represents the prediction error obtained by fitting the Gaussian mixture model;

[0084] A Gaussian mixture model is used to estimate the error distribution:

[0085] ,

[0086] Where:

[0087] represents the weight of the i-th Gaussian component; Mean ,variance Normal distribution of

[0088] The revised photovoltaic output forecast value obtained through Monte Carlo sampling is as follows:

[0089] ,

[0090] Where:

[0091] Represents the revised photovoltaic output forecast value at time The photovoltaic output power; Indicates at time Corrected forecast error.

[0092] The photovoltaic output forecast value is obtained according to the historical measured meteorological data, and the photovoltaic output curve is generated according to the photovoltaic output forecast value at different times and the sampled photovoltaic output forecast error.

[0093] The beneficial effect achieved by the present invention compared with the prior art lies in that, by analyzing the difference in the probability distribution of prediction errors when the photovoltaic power prediction values ​​are in different intervals through historical data, a Gaussian mixture model is used to fit the probability density distribution of the photovoltaic power prediction error; a distributed photovoltaic prediction error model is established to accurately characterize the distributed photovoltaic prediction error, track the fluctuation of distributed photovoltaic output, and obtain a distributed photovoltaic power output scenario that fits the actual operating status, which is conducive to characterizing the accurate voltage and loss distribution and facilitating the formulation of energy-saving and loss reduction strategies.

[0094] Step 2: According to the photovoltaic output curve and the monitoring value of the distribution network, the real-time flow state of energy in the distribution network is extracted to correct the photovoltaic output curve and adjust the charging and discharging strategy of the energy storage device;

[0095] Preferably, step 2 comprises:

[0096] According to the revised photovoltaic output curve , extract the characteristics of the photovoltaic output curve, including peak value, valley value and change rate;

[0097] Combining the photovoltaic output curve with real-time monitoring data, the energy flow is analyzed as follows:

[0098] ,

[0099] Where:

[0100] Indicates at time The charge and discharge power; Indicates at time load demand; Represents the energy loss caused by grid loss during transmission;

[0101] Analyze the real-time load demand of each node, monitor the voltage level of each node, and keep the voltage within the set range; calculate the node power balance as follows:

[0102] ,

[0103] Where:

[0104] Indicates Active power of each node; Indicates The voltage of each node; Indicates The current of each node; Indicates The power factor of each node;

[0105] According to real-time data, the photovoltaic output and the charging and discharging strategies of energy storage equipment are dynamically adjusted by setting rules; the setting rules are as follows:

[0106] When the predicted photovoltaic output When the load is higher than the demand, the energy storage is charged first;

[0107] When photovoltaic output When the power is lower than the load demand, discharge is given priority to supplement the load;

[0108] Particle swarm optimization is used to determine the best energy transmission path and maximize energy transmission efficiency by adjusting line switches and transformer configurations.

[0109] Preferably, the optimization algorithm used in the optimization calculation includes but is not limited to a genetic algorithm, a particle swarm optimization algorithm or a deep learning algorithm.

[0110] Particle swarm optimization is used to determine the optimal energy transmission path, as follows:

[0111] Determine the basic parameters of this optimization, select a certain number of particles, set the initial value of the variable to be optimized for each particle, and calculate the fitness of all particles;

[0112] According to the requirements, the fitness of all particles is analyzed as a whole, and the values ​​of the parameters to be optimized of all particles are re-determined to obtain new results;

[0113] For the particle that has completed the calculation first, directly assign a value to it and recalculate the path plan corresponding to the particle;

[0114] When the second particle completes the calculation, adjust its value to be close to the value of the first particle, and recalculate the corresponding path plan; assign values ​​to all particles in turn, analyze the results and determine whether the value is close to the value assigned by the previous particle; if it is close, assign the value directly; if not, recalculate the value until the value is close to the value assigned by the previous particle;

[0115] Repeat the above steps until the fitness of all particles meets the predetermined requirements, end the calculation and obtain the optimal energy transmission path.

[0116] Compared with the prior art, the significant difference between the present invention and the prior art is that the particle swarm optimization algorithm is used to realize data classification analysis, thereby actively realizing energy-saving and loss-reducing optimization operations on the distribution network;

[0117] Step 3: Based on the revised photovoltaic output curve and the adjusted charging and discharging strategy of the energy storage equipment, a multi-objective optimization model is constructed with the objectives of maximizing photovoltaic output, minimizing energy storage charging and discharging efficiency, and minimizing operating costs. When training the multi-objective optimization model, the weights corresponding to each objective are dynamically adjusted according to the actual operating parameters of the distribution network, the load demand forecast, and the energy supply change forecast.

[0118] Preferably, step 3 comprises:

[0119] According to the operation requirements and characteristics of the distribution network, the optimization objectives are determined to include: maximizing the self-use rate of photovoltaic output, minimizing the energy storage charging and discharging losses, and minimizing the operation costs;

[0120] The self-use rate of maximizing photovoltaic output is calculated as follows:

[0121] ,

[0122] Where:

[0123] Indicates the self-use rate of photovoltaic output; Indicates the power of photovoltaic output directly used for load; Indicates the total photovoltaic output power;

[0124] Minimize the energy storage charging and discharging loss by the following formula

[0125] ,

[0126] Where:

[0127] Indicates energy storage loss; Indicates time Charging power of energy storage; Indicates time Discharge power of energy storage; Indicates the energy storage charging efficiency; Indicates the energy storage discharge efficiency;

[0128] Minimize operating cost by calculating

[0129] ,

[0130] Where:

[0131] represents the total operating cost; Indicates time The power of purchased electricity; Indicates time the price of purchased electricity; Indicates time Equipment maintenance costs.

[0132] Preferably, the comprehensive objective function is calculated as follows:

[0133] ,

[0134] Where:

[0135] represents the comprehensive optimization objective; , and are the weight factors for each target respectively.

[0136] Step 4: Based on the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment generated by the trained multi-objective optimization model, the deviation between the actual photovoltaic output and the photovoltaic output scheduling plan is monitored in real time. According to the changes in the node load and voltage of the distribution network, the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment are updated to control the operation of the distribution network.

[0137] Preferably, step 4 comprises:

[0138] Through the multi-objective optimization model, the optimal operation strategy of the distribution network is obtained, and the optimization results are applied to the actual operation of the distribution network, including: adjusting the output power of photovoltaic power, the charging and discharging strategy of the energy storage system, and coordinating the purchase of electricity to balance the load demand and supply of the distribution network;

[0139] During the actual operation of the distribution network, the actual power output of the photovoltaic output is monitored in real time through the photovoltaic output curve, and the operation deviation of the distribution network is judged in combination with the load demand and energy storage status; according to the changes in the photovoltaic output curve, the operation strategy is corrected in real time, including adjusting the photovoltaic output power and the charging and discharging of energy storage; the operation deviation of the distribution network is adjusted through the feedback mechanism;

[0140] Through real-time monitoring and feedback, the self-learning closed loop automatically and gradually corrects the photovoltaic output curve and optimization strategy based on historical and real-time data;

[0141] When all optimization targets of the distribution network in actual operation deviate from the predicted values ​​within the set range, the self-learning closed loop completes the optimization task; in future operations, when new changes and anomalies are detected, the next round of optimization and adjustment will be automatically triggered.

[0142] Preferably, the operation deviation of the distribution network is adjusted through a feedback mechanism, including:

[0143] When there is a difference between the photovoltaic output curve and the predicted value, the charging and discharging behavior of the energy storage device is adjusted by analyzing the generated photovoltaic output prediction value and the corrected photovoltaic output curve to match the current power demand with the photovoltaic output curve;

[0144] When load demand changes, based on the analysis of photovoltaic output curve and load demand and voltage level of distribution network nodes, the energy dispatch of energy storage devices is dynamically optimized, and charging and discharging timing is reasonably adjusted to maintain energy flow balance at each node in the power grid;

[0145] According to the adjustment of future operation demand, the weight of each optimization target is re-evaluated through the multi-objective optimization model to adapt to the changes in distribution network operation and optimize the energy storage strategy.

[0146] Example 2 of the present invention provides an energy-saving and loss-reduction optimization system for a distributed photovoltaic active distribution network, including: a photovoltaic output scheduling module, an energy flow analysis module, a multi-objective optimization model module, and a real-time monitoring and feedback module;

[0147] The photovoltaic output dispatching module is used to uniformly dispatch distributed photovoltaic output through active coordination strategies, generate photovoltaic output forecast values ​​using historical meteorological data and real-time monitoring information, and correct the forecast error to form a photovoltaic output curve;

[0148] The energy flow analysis module is used to monitor the operating parameters of the distribution network in real time. Combined with the corrected photovoltaic output curve, it extracts the real-time flow status of energy in the distribution network. Based on the monitoring results, it dynamically adjusts the photovoltaic output and the charging and discharging strategy of the energy storage equipment to maintain the balance of energy supply and demand.

[0149] The multi-objective optimization model module is used to build and train a multi-objective optimization model based on the objectives of maximizing photovoltaic output, minimizing energy storage charging and discharging efficiency, and minimizing operating costs. During the training process, the weights of each optimization objective are dynamically adjusted according to the actual operating parameters of the distribution network, load demand forecast values, and energy supply change forecast values.

[0150] The real-time monitoring and feedback module is used to monitor the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment in real time, monitor the deviation between the actual photovoltaic output and the scheduling plan, and timely update the photovoltaic output scheduling plan and the charging and discharging strategy of the energy storage equipment according to the changes in the node load and voltage of the distribution network.

[0151] The beneficial effects of the present invention include at least: the present invention can realize energy-saving and loss-reduction optimization of the distribution network, improve the energy efficiency of the distribution network and reduce losses through the implementation of the distributed photovoltaic active distribution network energy-saving and loss-reduction optimization method and system. The present invention provides strong support for the sustainable development of the distribution network and has important practical application value.

[0152] In order to accurately track the fluctuation of distributed photovoltaic output, an optimization method for energy saving and loss reduction of active distribution network containing distributed photovoltaic is proposed. The difference in probability distribution of prediction error when photovoltaic power prediction value is in different ranges is analyzed through historical data, and the probability density distribution of photovoltaic power prediction error is fitted by Gaussian mixture model. A distributed photovoltaic prediction error model is established to accurately characterize the distributed photovoltaic prediction error, track the fluctuation of distributed photovoltaic output, and obtain the distributed photovoltaic power output scenario that fits the actual operating status, which is conducive to the accurate characterization of voltage and loss distribution and the formulation of energy saving and loss reduction strategies.

[0153] Taking into account the load type of the distribution system, a comprehensive load model for energy-saving and voltage-reduction control is established. Through the comprehensive load model for energy-saving and voltage-reduction control, the load consumption is expressed as a function of voltage, and the connection between load power and voltage is established. Under the premise of ensuring the quality of energy supply, the voltage of the distribution transformer feeder can be strategically reduced to reduce the overall energy consumption of the system, and energy saving and loss reduction can be achieved by voltage regulation.

[0154] The embodiments of the present invention disclose preferred embodiments, but are not limited thereto. A person skilled in the art can easily understand the spirit of the present invention based on the above embodiments and make different extensions and changes, but as long as they do not deviate from the spirit of the present invention, they are all within the protection scope of the present invention. The present disclosure may be a system, method and / or computer program product. The computer program product may include a computer-readable storage medium on which computer-readable program instructions for enabling a processor to implement various aspects of the present disclosure are loaded.

[0155] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing energy saving and loss reduction of distributed photovoltaic active distribution network, characterized in that: The following steps are involved: Through active coordination strategy, the distributed photovoltaic output is uniformly dispatched to generate photovoltaic output forecast value, and the error of photovoltaic output forecast value is corrected to form photovoltaic output curve; According to the photovoltaic output curve and the monitoring value of the distribution network, the real-time flow state of energy in the distribution network is extracted to correct the photovoltaic output curve and adjust the charging and discharging strategy of the energy storage device; including: Extract characteristic values ​​and corresponding time of characteristic values ​​on the photovoltaic output curve, where the characteristic values ​​include peak value, valley value and change rate; The distribution network monitoring values ​​at the time corresponding to the characteristic values ​​are used, where the distribution network monitoring values ​​include the energy storage power monitoring value, the node load demand monitoring value, and the grid loss monitoring value, to extract the real-time flow state of energy in the distribution network as follows: In the formula, Indicates that the eigenvalue corresponds to time Energy storage power monitoring value under Indicates that the eigenvalue corresponds to time The node load demand monitoring value under Indicates that the eigenvalue corresponds to time The power grid loss monitoring value under is the characteristic value; According to the real-time load demand of each node, the voltage level of each node is monitored and adjusted when the voltage exceeds the set range. According to the real-time flow state of energy, the charging and discharging strategy of photovoltaic output and energy storage equipment is dynamically adjusted through the energy balance principle; the setting rules are as follows: When the photovoltaic output is greater than the node load demand, the energy storage device is charged first; When the photovoltaic output is less than the node load demand, the energy storage device discharges first to supplement the load; According to the revised photovoltaic output curve and the adjusted charging and discharging strategy of the energy storage equipment, a multi-objective optimization model is constructed with the goals of maximizing photovoltaic output, minimizing energy storage charging and discharging efficiency, and minimizing operating costs. When training the multi-objective optimization model, the weights corresponding to each objective are dynamically adjusted according to the actual operating parameters of the distribution network, the load demand forecast value, and the energy supply change forecast value; including: The self-use rate of maximizing photovoltaic output is calculated as follows: In the formula, Indicates the self-use rate of photovoltaic power output; Indicates the power of photovoltaic output directly used for load; Indicates the total photovoltaic output power; The minimum energy storage charging and discharging loss is calculated as follows: In the formula, Indicates energy storage loss; Indicates time Charging power of energy storage; Indicates time Discharge power of energy storage; Indicates the energy storage charging efficiency; Indicates the energy storage discharge efficiency; The minimum operating cost is calculated as follows: In the formula, represents the total operating cost; Indicates time The power of purchased electricity; Indicates time the price of purchased electricity; Indicates time Cost of equipment maintenance; The comprehensive objective function is calculated as follows: Where: represents the comprehensive optimization objective; , and are the weight factors for each target respectively; Based on the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment generated by the trained multi-objective optimization model, the deviation between the actual photovoltaic output and the photovoltaic output scheduling plan is monitored in real time. According to the changes in the node load and voltage of the distribution network, the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment are updated to control the operation of the distribution network.

2. The method for optimizing energy saving and loss reduction of a distributed photovoltaic active distribution network according to claim 1, characterized in that: Through the active coordination strategy, the distributed photovoltaic output is uniformly dispatched, the photovoltaic output forecast value is generated, and the photovoltaic output forecast value error is corrected to form a photovoltaic output curve, including: Collect actual output data of distributed photovoltaics, operating status data of distribution networks, and user load demand data, and generate photovoltaic output forecast values ​​based on historical meteorological data; Obtain historical photovoltaic output data; use statistical methods to fit the errors of historical photovoltaic output data, and correct the photovoltaic output prediction value based on the fitted errors.

3. The energy-saving and loss-reduction optimization method for a distributed photovoltaic active distribution network according to claim 2 is characterized in that: Obtain historical photovoltaic output data; use statistical methods to fit the errors of historical photovoltaic output data, and correct the photovoltaic output forecast value based on the fitted errors, including: The probability distribution of the error of the historical photovoltaic output data in different power ranges is fitted based on the Gaussian mixture model, and the error of the photovoltaic output prediction value is obtained from the probability distribution of the error using the Monte Carlo sampling method. According to the time sequence, the sum of the photovoltaic output prediction value and the photovoltaic output prediction value is used as the corrected photovoltaic output to form a photovoltaic output curve.

4. The energy-saving and loss-reduction optimization method for a distributed photovoltaic active distribution network according to claim 1, characterized in that: Dynamically adjust the weights of each target, including: In the multi-objective optimization model, each particle is regarded as a different weight allocation scheme; each particle represents a weight configuration, and the fitness is evaluated by the objective functions including maximizing photovoltaic output, minimizing energy storage efficiency, and minimizing operating costs; Construct a fitness function, evaluate the fitness value of each particle through the specific parameters of photovoltaic output, energy storage efficiency and operating cost, and obtain the optimal weight combination based on the actual operating conditions to balance the three objectives; Each particle adjusts the weight distribution through its own historical optimal position and global optimal position, and dynamically adjusts the weight configuration of the multi-objective optimization model; through the iterative optimization of the particle swarm algorithm, the best weight distribution scheme is gradually obtained; After the particle swarm algorithm converges to the global optimal solution, the final weight configuration is determined, and the optimization process of photovoltaic output, energy storage charging and discharging strategy and operating cost is adjusted based on the final weight configuration to achieve dynamic adjustment of the weights corresponding to each target.

5. The method for optimizing energy saving and loss reduction of a distributed photovoltaic active distribution network according to claim 1, characterized in that: Based on the photovoltaic output dispatching plan and the charging and discharging dispatching plan of the energy storage equipment generated by the trained multi-objective optimization model, the deviation between the actual photovoltaic output and the photovoltaic output dispatching plan is monitored in real time. According to the changes in the node load and voltage of the distribution network, the photovoltaic output dispatching plan and the charging and discharging dispatching plan of the energy storage equipment are updated to control the operation of the distribution network, including: Through the multi-objective optimization model, the optimal operation strategy of the distribution network is obtained, and the optimization results are applied to the actual operation of the distribution network, including: adjusting the power of photovoltaic output, optimizing the charging and discharging strategy of the energy storage system, and coordinating the purchase of electricity to achieve the balance between the load demand and power supply of each node in the distribution network; During the actual operation of the distribution network, the actual power output of the photovoltaic output is monitored in real time through the photovoltaic output curve, and the operation deviation of the distribution network is judged in combination with the node load demand and the status of the energy storage equipment; the operation strategy is corrected in real time according to the changes in the photovoltaic output curve; For the detected distribution network operation deviation, automatic adjustment is carried out through the built-in feedback mechanism. Based on historical data and real-time monitoring data, the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment are gradually corrected through a self-learning closed loop; When all optimization targets of the distribution network in actual operation deviate from the predicted values ​​within the set range, the self-learning closed loop completes the optimization task; in future operations, when new changes and anomalies are detected, the next round of optimization and adjustment will be automatically triggered.

6. The method for optimizing energy saving and loss reduction of a distributed photovoltaic active distribution network according to claim 5, characterized in that: The operation deviation of the distribution network is adjusted through feedback mechanisms, including: When there is a difference between the photovoltaic output curve and the predicted value, the difference between the generated photovoltaic output prediction value and the actual corrected photovoltaic output curve is analyzed, and the charging and discharging behavior of the energy storage device is adjusted to match the current power demand with the photovoltaic output curve; When the node load demand in the distribution network changes, based on the analysis of the photovoltaic output curve and the node load demand and voltage level of the distribution network, the energy dispatch of the energy storage device is dynamically optimized, and the charging and discharging timing is reasonably adjusted to maintain the energy flow balance of each node in the power grid; According to the adjustment of future operating needs, the weight of each optimization target is re-evaluated through the multi-objective optimization model to adapt to the changes in distribution network operation. Through weight adjustment, the dispatch strategy of energy storage equipment is optimized under different operating conditions.

7. A distributed photovoltaic active distribution network energy saving and loss reduction optimization system, comprising: Photovoltaic output dispatch module, energy flow analysis module, multi-objective optimization model module, and real-time monitoring and feedback module; The method for optimizing energy saving and loss reduction of a distributed photovoltaic active distribution network according to any one of claims 1 to 6 is characterized in that: The photovoltaic output dispatching module is used to uniformly dispatch distributed photovoltaic output through active coordination strategies, generate photovoltaic output forecast values ​​using historical meteorological data and real-time monitoring information, and correct the forecast error to form a photovoltaic output curve; The energy flow analysis module is used to monitor the operating parameters of the distribution network in real time. Combined with the corrected photovoltaic output curve, it extracts the real-time flow status of energy in the distribution network. Based on the monitoring results, it dynamically adjusts the photovoltaic output and the charging and discharging strategy of the energy storage equipment to maintain the balance of energy supply and demand. The multi-objective optimization model module is used to construct and train a multi-objective optimization model based on the objectives of maximizing photovoltaic output, minimizing energy storage charging and discharging efficiency, and minimizing operating costs. During the training process, the weights of various optimization objectives are dynamically adjusted according to the actual operating parameters of the distribution network, the load demand forecast value, and the energy supply change forecast value. Among them: by calculating the photovoltaic output self-use rate, the proportion of photovoltaic output directly used for the load is determined; the charging and discharging losses of the energy storage equipment are calculated, the losses of the energy storage system in the energy conversion process are evaluated, and the charging and discharging strategies of the energy storage equipment are optimized; the operating costs, including the cost of purchased electricity and the equipment maintenance costs, are calculated, and the photovoltaic output scheduling plan and the charging and discharging strategies of the energy storage equipment are optimized; combined with the above optimization objectives, comprehensive optimization is performed according to the set weight factors to achieve a balance between the photovoltaic output self-use rate, energy storage losses, and operating costs, and the photovoltaic output scheduling plan and the charging and discharging strategies of the energy storage equipment are generated; The real-time monitoring and feedback module is used to monitor the photovoltaic output scheduling plan and the charging and discharging scheduling plan of the energy storage equipment in real time, monitor the deviation between the actual photovoltaic output and the scheduling plan, and timely update the photovoltaic output scheduling plan and the charging and discharging strategy of the energy storage equipment according to the changes in the node load and voltage of the distribution network.

Images