Photovoltaic-energy storage-based energy storage site selection and power distribution network cooperative loss reduction optimization method

Abstract

The invention discloses a photovoltaic-energy storage-based energy storage site selection and power distribution network cooperative loss reduction optimization method, and the method comprises the steps: introducing a voltage-current coupling coefficient and a temperature correction factor, and constructing an improved line loss calculation model; calculating a photovoltaic access line loss sensitivity factor of each node, identifying a key node sensitive to line loss, and taking the key node as a site of an energy storage system ESS; based on the improved line loss calculation model and the site selection of the energy storage system ESS, establishing a self-adaptive line loss minimization control strategy, and combining a line loss-voltage joint optimization objective function to comprehensively balance a line loss expected value, a variance and voltage stability; a differential dynamic programming method is adopted to optimize an energy storage system ESS charging and discharging curve, and a high-dimensional alternating direction multiplier method is combined for distributed solution, so that cooperative loss reduction of multiple energy storage systems and a power distribution network is realized. According to the method, efficient loss reduction and stable operation of the high-permeability photovoltaic power distribution network are achieved, and meanwhile the economical efficiency and the renewable energy consumption capacity of the power distribution network are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of power distribution network optimization technology, and in particular to a method for optimizing energy storage site selection and power distribution network loss reduction based on photovoltaic-energy storage. Background Technology

[0002] In recent years, global warming, environmental degradation, and increased energy consumption have become a focus of attention for countries worldwide. Energy conservation, environmental protection, and flexible power generation technologies are some of the advantages of distributed generation, providing consumers with reliable, high-quality electricity. Distributed photovoltaic (PV) power generation is one of the most promising technologies in distributed generation due to its advantages such as abundant power reserves, security, and zero pollution.

[0003] After distributed photovoltaic (PV) systems are integrated into the power distribution network, their intermittent and fluctuating output characteristics introduce significant uncertainties to the network's operation. When large-scale distributed PV systems are integrated, the power flow in the distribution network may exhibit stochastic characteristics, including reverse or bidirectional power flow constraints. This leads to problems such as high power losses in the distribution network. Power losses are a key factor in distribution network planning and economic operation control. Therefore, optimizing the integration of PV systems and mitigating their adverse effects on the power system has become an extremely important problem that must be solved.

[0004] One effective method to reduce distribution network line losses is to integrate energy storage systems (ESS) into the distribution network. Since power loss is a quadratic function of current, using ESS to transfer current during peak loads or photovoltaic output can help reduce losses. For new power systems, especially those relying on renewable energy, ESS can effectively reduce line losses in the distribution network. However, the location and capacity determination of the ESS have a significant impact on its effectiveness in reducing losses. Furthermore, coordinated optimization of distributed photovoltaic systems and ESS can achieve peak shaving and valley filling, thereby effectively reducing line losses in the distribution network. Therefore, this invention proposes a photovoltaic-energy storage-based energy storage location and distribution network coordinated loss reduction model and optimization method. Summary of the Invention

[0005] The purpose of this invention is to provide a photovoltaic-energy storage-based energy storage site selection and distribution network loss reduction optimization method. This method significantly improves the economy of the distribution network and the renewable energy absorption capacity while ensuring voltage stability.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] A photovoltaic-energy storage-based method for optimizing energy storage site selection and distribution network loss reduction includes the following steps:

[0008] S1. By introducing voltage-current coupling coefficient and temperature correction factor, an improved line loss calculation model is constructed to quantify the nonlinear effects of power factor and the dynamic characteristics of line resistance with temperature.

[0009] S2. Calculate the photovoltaic access line loss sensitivity factor of each node, identify the key nodes that are sensitive to line loss, and use them as the site selection for the energy storage system (ESS).

[0010] S3. Based on the improved line loss calculation model and the location of the energy storage system ESS, an adaptive line loss minimization control strategy is established. Combined with the line loss-voltage joint optimization objective function, the expected value of line loss, variance and voltage stability are comprehensively weighed.

[0011] S4. The differential dynamic programming method is used to optimize the ESS charge and discharge curve of the energy storage system. Combined with the high-dimensional alternating direction multiplier method for distributed solution, the collaborative loss reduction of multiple energy storage systems and distribution network is realized.

[0012] Preferably, in S1, the improved line loss calculation model is as follows:

[0013]

[0014] Wherein, ΔP Loss_m1 R is the line loss between node m-1 and node m. m (T) represents the line resistance between node m-1 and node m at temperature T, U m P is the effective value of the voltage at node m. j Let Q be the active power load of node j. j For the reactive power load at node j, P pv Q represents the active power output of the photovoltaic system. pv For the reactive power output of the photovoltaic system, These represent the active power during charging, active power during discharging, reactive power during charging, and reactive power during discharging of the energy storage system at time t, respectively, ω. t For time-varying weighting coefficients, SOC ess For the energy storage system, f(SOC) represents the state of charge. ess ) is a function of the energy storage state of charge.

[0015] Preferably, in step S2, calculating the photovoltaic access line loss sensitivity factor for each node and identifying key nodes sensitive to line loss specifically includes:

[0016] After the distributed photovoltaic system is connected to the power distribution network, the active power and reactive power of the photovoltaic access node are corrected.

[0017] Based on the total line resistance from the node with load to the root node, the sensitivity of power change to line loss is quantified to obtain the photovoltaic access line loss sensitivity factor.

[0018] Select the node with the highest average value of photovoltaic access line loss sensitivity factor within 24 hours to deploy ESS.

[0019] Preferably, S2 also includes introducing a line loss sensitivity factor for load fluctuations to suit different load characteristics:

[0020]

[0021] Wherein, ΔP Loss ΔL represents the change in line loss. i The load change at node i, × i σ(L) is the weighting coefficient for node positions. i Let L be the standard deviation of the load at node i. i Let be the load power of node i.

[0022] Preferably, in S3, the objective function of the adaptive line loss minimization control strategy is as follows:

[0023]

[0024] Where minJ is the objective function, t is the time series index, T is the total number of time intervals within the optimization period, and λ t μ is the weighting coefficient for the expected value of line loss at time t. t ν is the weighting coefficient for the variance of line loss at time t. t Let E{ΔP} be the weighting coefficient of the volatility sensitivity factor at time t. Loss,t Let} be the expected value of the line loss at time t, Var{ΔP} Loss,t Let VSF be the variance of the line loss at time t. t Let be the volatility sensitivity factor at time t.

[0025] Preferably, the objective function for the joint optimization of line loss and voltage is:

[0026]

[0027] Where α is the line loss weighting coefficient, β is the mean square voltage deviation weighting coefficient, γ is the maximum voltage deviation weighting coefficient, and ΔP Loss,0 For baseline loss, U ref The reference voltage value is N, and the total number of nodes in the distribution network is N.

[0028] Preferably, in S4, the optimization of the ESS charge-discharge curve using a differential dynamic programming method includes:

[0029] Define the value function V(x) T ) and terminal cost function φ(x T The control variables are updated iteratively using incremental dynamic programming, as shown in the following formula:

[0030]

[0031] in, This represents the update amount of the control variable in the (i+1)th iteration. Let be the second derivative of the Hamiltonian function with respect to the control variable. The first derivative of the Hamiltonian function with respect to the control variable is... This is the mixed second derivative of the Hamiltonian function with respect to the control and state variables. Update the state variables for the (i+1)th iteration.

[0032] Preferably, in S4, the distributed solution combining the high-dimensional alternating direction multiplier method includes:

[0033] The global optimization problem is decomposed into multiple subproblems, and a distributed solution using the high-dimensional alternating direction multiplier method is employed.

[0034]

[0035] Where x represents the voltage adjustment and reactive power compensation at each node, z represents the charging and discharging power of the energy storage system, and y represents the Lagrange multiplier corresponding to the power balance constraint; collaborative optimization is achieved by augmenting the Lagrange function. for:

[0036]

[0037] Where f(x) is the line loss objective function related to grid voltage and reactive power, g(z) is the cost and loss function related to energy storage system charging and discharging, Ax is the power equivalent of grid node regulation, Bz is the power injection amount of ESS charging and discharging, c is the baseline value of system net power demand, and y T Let y be the dual variable of y, and ρ be the penalty parameter.

[0038] The present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, wherein when the computer program is executed by a processor, it implements an optimization method for energy storage site selection and distribution network collaborative loss reduction based on photovoltaic-energy storage as described above.

[0039] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0040] (1) The photovoltaic-energy storage-based energy storage location and distribution network collaborative loss reduction method proposed in this invention achieves efficient loss reduction and stable operation of high-penetration photovoltaic distribution networks by improving the line loss model (including voltage-current coupling and temperature correction), photovoltaic sensitivity factor (PAVSF) location, adaptive multi-objective control strategy (line loss-voltage joint optimization) and DDP-ADMM collaborative algorithm.

[0041] (2) The improved model of this invention reduces the error in line loss calculation. The ESS layout combined with PAVSF site selection reduces line loss compared to random layout. Through adaptive weight adjustment and DDP-ADMM distributed solution, dynamic optimization of energy storage charging and discharging and voltage regulation is achieved, improving peak shaving and valley filling efficiency. The robustness is enhanced. The variance control term and temperature correction factor effectively suppress the influence of photovoltaic fluctuations and resistance temperature changes, reducing voltage deviation. The ADMM distributed framework decomposes the global optimization problem into parallel sub-problems, improving the solution speed compared to the traditional centralized method. Overall, this method significantly improves the economy of the distribution network and the renewable energy absorption capacity while ensuring voltage stability. Attached Figure Description

[0042] 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 introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 A flowchart illustrating the photovoltaic-energy storage-based energy storage location selection and distribution network collaborative loss reduction optimization method provided by this invention;

[0044] Figure 2 This is a network diagram of the IEEE 33 power distribution system in an embodiment of the present invention;

[0045] Figure 3 The PAVSF value of the IEEE 33-node distribution network in this embodiment of the invention;

[0046] Figure 4 This is a diagram showing the optimal energy storage capacity in an embodiment of the present invention;

[0047] Figure 5 This is a diagram showing the total system power loss over 24 hours in an embodiment of the present invention;

[0048] Figure 6 This is a bus loss diagram of the 24-hour branch circuit in an embodiment of the present invention;

[0049] Figure 7 This is a comparison diagram of different combinations of energy storage locations in embodiments of the present invention. Detailed Implementation

[0050] 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, and 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.

[0051] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0052] like Figure 1 As shown, the present invention provides a method for optimizing energy storage site selection and distribution network loss reduction based on photovoltaic-energy storage, comprising the following steps:

[0053] S1. By introducing voltage-current coupling coefficient and temperature correction factor, an improved line loss calculation model is constructed to quantify the nonlinear effects of power factor and the dynamic characteristics of line resistance with temperature.

[0054] S2. Calculate the photovoltaic access line loss sensitivity factor of each node, identify the key nodes that are sensitive to line loss, and use them as the site selection for the energy storage system (ESS).

[0055] S3. Based on the improved line loss calculation model and the location of the energy storage system ESS, an adaptive line loss minimization control strategy is established. Combined with the line loss-voltage joint optimization objective function, the expected value of line loss, variance and voltage stability are comprehensively weighed.

[0056] S4. The differential dynamic programming method is used to optimize the ESS charge and discharge curve of the energy storage system. Combined with the high-dimensional alternating direction multiplier method for distributed solution, the collaborative loss reduction of multiple energy storage systems and distribution network is realized.

[0057] Existing line loss calculation methods mainly consider the relationship between power and voltage, but neglect the nonlinear coupling effect between voltage and current. In a specific embodiment, this invention introduces a voltage-current coupling coefficient, and the line loss calculation model incorporating the voltage-current coupling effect is as follows:

[0058]

[0059] in, The line loss between node i-1 and node i, in kW

[0060] R i The line resistance between node i-1 and node i, in Ω;

[0061] U i: The effective value of the voltage at node i, in kV;

[0062] p j : Active power load at node j, in kW;

[0063] Q j : Reactive power load at node j, in kVar;

[0064] Z: End node number of the power distribution line, indicating the total number of nodes in the line;

[0065] α i Voltage-current coupling coefficient, dimensionless, typically ranging from 0.01 to 0.05, reflects the influence of power factor on line loss.

[0066] Line resistance varies with temperature, which significantly impacts line loss calculations. An adaptive line loss calculation method considering the temperature characteristics of line impedance is proposed. A temperature correction factor is introduced.

[0067] R i (T)=R i (T0)·[1+β·(T-T0)]

[0068] Among them, R i (T): Line resistance at the current temperature T, in Ω;

[0069] R i (T0): Line resistance at reference temperature T0, in Ω;

[0070] β: Temperature coefficient of conductor material, approximately 0.00393 / °C for copper conductor and approximately 0.00403 / °C for aluminum conductor; T: Current conductor temperature, in °C; T0: Reference temperature, usually taken as 20°C.

[0071] When photovoltaic and energy storage facilities are simultaneously connected to the distribution network, the comprehensive line loss calculation formula under multi-energy complementarity can be expressed as:

[0072]

[0073] Wherein, ΔP Loss_m1 Line loss between node m-1 and node m, in kW;

[0074] R m (T): Line resistance between node m-1 and node m at temperature T, in Ω;

[0075] U m : The effective value of the voltage at node m, in kV;

[0076] P pvActive power output of a photovoltaic system, measured in kW;

[0077] Q pv Reactive power output of a photovoltaic system, measured in kVar;

[0078] The charging active power, discharging active power, charging reactive power, and discharging reactive power of the energy storage system at time t;

[0079] ω t Time-varying weighting coefficients, dimensionless, used to adjust the degree of influence of energy management strategies;

[0080] SOC ess The state of charge of the energy storage system, with a value ranging from 0 to 1;

[0081] f(SOC ess ): A function of the energy storage state of charge, reflecting the impact of energy management strategies on line losses, in kW.

[0082] For different load characteristics, a line loss sensitivity factor (VSF) based on load volatility is introduced:

[0083]

[0084] Wherein, ΔP Loss Line loss variation, in kW;

[0085] ΔL i : The load change at node i, in kW;

[0086] γ i : Node location weighting coefficient, dimensionless, the value decreases as it gets closer to the substation;

[0087] σ(L i ): Standard deviation of the load at node i, in kW, used to characterize load variability;

[0088] L i : Load power of node i, in kW.

[0089] The line loss sensitivity factor (VSF) measures the impact of load changes at a specific node on the total power loss of a distribution network. This invention proposes a method based on the photovoltaic access line loss sensitivity factor (PAVSF) for selecting the location of an energy storage system (ESS). By calculating the PAVSF value of each node, the critical nodes most sensitive to line losses can be identified, thus providing a basis for ESS deployment. After the distributed photovoltaic system is connected to the distribution network, the active and reactive power at the nodes are first corrected according to the following formula:

[0090]

[0091] Assuming that only one node i has a load in the distribution network, the total power loss from the root node s to node i can be expressed as:

[0092]

[0093] In the formula, ΔP si P represents the total power loss from root node s to node i. i 'Q' represents the active power at node i. i Represents the reactive power at node i, ∑R si This represents the total resistance of the line from root node s to node i.

[0094] Similarly, the active power loss ΔP from node i to root node s is It can be represented as:

[0095]

[0096] From ΔP si =ΔP is We can obtain:

[0097]

[0098] For ΔP si Regarding ∑R si The partial derivatives are calculated and set to zero to minimize the loss, as shown below:

[0099]

[0100] To quantify the impact of power variation at node i on losses, this invention defines a photovoltaic line loss sensitivity factor PF. i for:

[0101]

[0102] Calculate the hourly PAVSF value using the method described above, and then calculate the average of all PAVSF values ​​over 24 hours to obtain the daily PAVSF value.

[0103] The integration of distributed photovoltaic (PV) power generation into the distribution network involves uncertainty and randomness. The PAVSF value represents the sensitivity of the randomness of distributed PV power generation integration to the losses of the corresponding nodes. Therefore, these values ​​can be used to identify key nodes suitable for installing energy storage systems.

[0104] Based on the above-mentioned energy storage site selection and improved line loss calculation method, an adaptive line loss minimization control strategy is proposed:

[0105]

[0106] Wherein, minJ: the objective function, is the comprehensive evaluation index that needs to be minimized;

[0107] t: Time series index, t = 1, 2, ..., T;

[0108] T: The total number of time intervals within the optimization period;

[0109] λ t : Weighting coefficient for the expected value of line loss at time t, dimensionless;

[0110] μ t : The weighting coefficient of the variance of line loss at time t, dimensionless;

[0111] v t : The weighting coefficient of the volatility sensitivity factor at time t, dimensionless;

[0112] E{ΔP Loss,t}: The expected value of line loss at time t, in kW;

[0113] Var{ΔP Loss,t}: The variance of line loss at time t, in kW²;

[0114] VSF t : Volatility sensitivity factor at time t, in units of kW / (kW·kW).

[0115] Existing methods often consider line loss and voltage optimization separately. This method proposes a joint line loss-voltage optimization objective function:

[0116]

[0117] Where α: line loss weighting coefficient, dimensionless;

[0118] β: Weighting coefficient for the mean square value of voltage deviation, dimensionless;

[0119] γ: Maximum voltage deviation weighting coefficient, dimensionless;

[0120] ΔP Loss,0 Baseline line loss (usually the line loss before optimization), in kW;

[0121] U ref Reference voltage value, in kV;

[0122] N: The total number of nodes in the distribution network.

[0123] To reduce line losses in distribution networks more accurately and dynamically, a multi-level, multi-scale optimization solution method for distribution networks is further proposed.

[0124] To optimize the charge-discharge curves of energy storage systems, a differential dynamic programming (DDP) method is employed.

[0125]

[0126] Where V(x) T Terminal status x T Value function;

[0127] φ(x T Terminal cost function;

[0128] V(x t State x at time t t Value function;

[0129] u t : Control variables at time t (such as energy storage charging and discharging power);

[0130] L(x t ,u t ): Cost function for the current step;

[0131] V(f(x t ,u t ): System state transition function, describing the transition from x t To x t+1 Dynamic characteristics;

[0132] Incremental dynamic programming iteration:

[0133]

[0134] The amount of control variable update in the (i+1)th iteration;

[0135] The second derivative of the Hamiltonian function with respect to the control variables (Hessian matrix);

[0136] The first derivative (gradient vector) of the Hamiltonian function with respect to the control variable;

[0137] The mixed second derivative of the Hamiltonian function with respect to control and state variables;

[0138] Update of state variables in the (i+1)th iteration;

[0139] The global optimization problem is decomposed into multiple subproblems and solved in a distributed manner using the high-dimensional alternating direction multiplier method (ADMM):

[0140]

[0141] Where x: voltage adjustment and reactive power compensation at each node;

[0142] z: Charging and discharging power of the energy storage system;

[0143] y: Lagrange multipliers corresponding to power balance constraints;

[0144] Augmented Lagrange function for:

[0145]

[0146] Where f(x) is the line loss objective function related to grid voltage and reactive power, g(z) is the cost and loss function related to energy storage system charging and discharging, Ax is the power equivalent of grid node regulation, Bz is the power injection amount of ESS charging and discharging, c is the baseline value of system net power demand, and y T Let y be the dual variable, and ρ be the penalty parameter used to adjust the degree of penalty for constraint violation.

[0147] The ADMM distributed solution method enables efficient calculation of line loss-voltage co-optimization of multiple energy storage systems and distribution networks, overcoming the problems of high computational complexity and heavy communication burden of traditional centralized solution methods.

[0148] To minimize line loss, a novel iterative update mechanism is proposed. The core iterative update formula can be expressed as a multidimensional composite function:

[0149]

[0150] in, The update representing particle velocity not only includes key terms from the traditional PSO algorithm but also introduces several adjustment mechanisms. ω is the inertia weight, controlling the particle's historical motion inertia; c1 and c2 are learning factors, balancing local optima and global optima; r1 and r2 are random factors, increasing the algorithm's randomness.

[0151] In the iterative update formula, three innovative adjustment terms are introduced: δ·η(t), λ·ξ(x) i ) and μ·ζ(x i ).

[0152]

[0153] η(t) is a time-dependent sinusoidal decaying function inspired by periodic variations in natural systems. This function not only introduces a time dimension but also simulates the decrease in exploration intensity during system convergence through an exponential decay term.

[0154]

[0155] ξ(x i The fitness distribution is a nonlinear mapping function. It dynamically adjusts the particle's search strategy by comparing the fitness of a single solution with the overall fitness distribution. When the fitness of a solution deviates significantly from the average level, it significantly affects the particle's velocity update. This design effectively balances the algorithm's global search capability and local convergence capability.

[0156]

[0157] Furthermore, ζ(x) was introduced. i This is a logistic transformation function based on population diversity. It dynamically adjusts the exploration intensity of the algorithm by measuring the diversity of solutions in the population. When population diversity decreases, ζ(x) i This increases the randomness of the algorithm and prevents premature convergence.

[0158] This multi-dimensional, adaptive performance evaluation method enables the algorithm to overcome the limitations of traditional optimization methods. It not only pursues the optimization of a single metric, but also the synergistic optimization of the entire system.

[0159] This invention proposes a photovoltaic-energy storage system location and distribution network collaborative loss reduction model and optimization method. First, by introducing a voltage-current coupling coefficient and a temperature correction factor, the traditional line loss calculation model is improved, addressing the nonlinear impact of line resistance variations with temperature and power factor on losses. Second, a photovoltaic line loss sensitivity factor (PAVSF) is defined to quantify the sensitivity of each node to line losses after photovoltaic integration, providing a scientific basis for the optimal location of the energy storage system. Based on this, an adaptive line loss minimization control strategy is constructed, combining a line loss-voltage joint optimization objective function to comprehensively balance the expected value of line loss, variance, and voltage stability. Furthermore, differential dynamic programming (DDP) is used to optimize the energy storage charge-discharge curve, and the high-dimensional alternating direction multiplier method (ADMM) is used to achieve distributed solution, significantly reducing computational complexity. An innovative multi-dimensional adaptive optimization mechanism is introduced, balancing global search and local convergence capabilities through a periodic decay function, nonlinear fitness mapping, and a population diversity adjustment term. This method effectively improves the loss reduction efficiency and robustness of distribution networks in high-penetration photovoltaic scenarios through multi-level collaborative optimization.

[0160] To verify the effectiveness of the photovoltaic-energy storage-based energy storage location and distribution network collaborative loss reduction model and optimization method in the distribution network, a simulation experiment was conducted on the IEEE 33 bus system.

[0161] 1. Results Analysis

[0162] Networks in simulation systems, such as Figure 2As shown, the system parameters are based on the standard IEEE 33 bus test case, with a voltage reference value of 10kV. Distributed photovoltaic (PV) systems are integrated at nodes 2, 4, 8, 14, 18, 24, 30, and 32 to simulate the state of high-penetration distributed PV integration into the distribution network. Through multidimensional mathematical analysis of the IEEE 33-node system, the loss-sensitivity PAVSF factor of each node was calculated. The results show that the sensitivity factor of the nodes varies significantly under different load and generation conditions. Specific results are as follows... Figure 3 As shown.

[0163] The PAVSF calculation method introduced in the ESS location method based on the photovoltaic access loss sensitivity factor can be used to determine the PAVSF values ​​of all nodes after distributed photovoltaic access to the IEEE 33 system. The results are as follows: Figure 2 As shown.

[0164] from Figure 3 In the analysis, eight nodes with the highest PAVSF values ​​were identified in the IEEE 33-node system: nodes 24, 31, 8, 25, 32, 30, 14, and 18. These nodes exhibited significant sensitivity under different distributed photovoltaic (PV) integration, load variations, and power generation conditions, indicating their substantial impact on the overall line losses of the distribution network. Therefore, identical energy storage systems (ESS) were installed on these eight selected nodes to analyze the effectiveness of the PV-energy storage synergy approach in reducing distribution network losses.

[0165] like Figure 4 As shown, the system line loss curve exhibits a quadratic function pattern as the access capacity of the ESS gradually increases. With the increase in ESS capacity, the system line loss first decreases and then increases again. Therefore, the optimal access capacity of the ESS occurs when the system line loss reaches its minimum.

[0166] First, simulation analysis was conducted to reduce losses through photovoltaic-energy storage coordination for different types of residential loads. At midday, electricity demand is relatively low, while the output of the photovoltaic system is excessive, leading to excess photovoltaic power generation and thus increasing, rather than decreasing, system line losses. After integrating an energy storage system (ESS), the excess electricity generated by distributed photovoltaics during midday is stored in the grid-side ESS. Around 8 PM, when electricity demand is high and distributed photovoltaic output is low, unable to meet user load demands, the energy stored in the ESS is released for user use, thereby reducing system line losses. This experiment included three control groups: one without distributed photovoltaics and ESS, one with only distributed photovoltaics, and one with both distributed photovoltaics and ESS. The results obtained include the total active power loss of the system over 24 hours and the total active power loss of each branch over 24 hours, such as... Figures 5-6 As shown.

[0167] like Figure 5 As shown, when a distributed photovoltaic system is connected to a grid-side energy storage system (ESS) simultaneously, the line losses of the residential load system are significantly reduced during the midday and evening periods. Figure 6 As shown, after the simultaneous integration of distributed photovoltaic and ESS, the total line loss of the system branch over 24 hours is significantly reduced, achieving effective loss reduction results.

[0168] Based on this, different ESS layout combinations were simulated and analyzed. The ESS layout combinations based on PAVSF were compared with other random ESS layout combinations, and the results are as follows: Figure 7 As shown.

[0169] Scenario 1 represents a PAVSF-based ESS location combination, while other scenarios represent random ESS location combinations. For example... Figure 7 As shown, the ESS arrangement based on PAVSF achieves the best loss reduction effect, verifying the feasibility of the proposed method.

[0170] 2. Conclusion

[0171] The experimental portion of this invention is based on the IEEE 33-node system. Simulations validate the effectiveness of the photovoltaic-energy storage synergistic loss reduction method based on the Photovoltaic Access Loss Sensitivity Factor (PAVSF). The experiment is designed in three stages: First, the eight nodes most sensitive to line loss (nodes 24, 31, 8, 25, 32, 30, 14, and 18) are identified using PAVSF calculations, and an energy storage system (ESS) is deployed based on these nodes. Second, the correlation between ESS capacity and system line loss is analyzed, revealing that line loss changes quadratically with increasing capacity, and that there exists an optimal capacity point that minimizes line loss. Finally, the loss reduction effects of PAVSF-based ESS deployment and random deployment are compared.

[0172] During the midday peak of photovoltaic (PV) output and the evening peak of load demand, ESS significantly reduced line losses through peak shaving and valley filling. Compared to a control group without PV-storage, the PAVSF-based synergistic solution reduced total 24-hour line losses by approximately 23%. The ESS capacity optimization curve shows that when the capacity increases to a critical value, the system line loss reaches its lowest point. Figure 4 This verifies the crucial role of capacity configuration in loss reduction. The PAVSF-based ESS layout (Scenario 1) achieves superior loss reduction at the same capacity compared to random layouts (other scenarios). Figure 7 The reduction in line loss was increased by about 15%, highlighting the scientific nature and reliability of the PAVSF site selection strategy.

[0173] In addition, the experiment analyzed the branch-level line loss ( Figure 6The results show that the synergistic optimization of photovoltaic and energy storage can balance the power distribution of each branch of the distribution network, avoid local overload, and further improve the overall stability of the system. The experimental results comprehensively confirm that the proposed method has significant loss reduction capabilities and engineering application potential in high-penetration photovoltaic scenarios.

[0174] The present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, wherein when the computer program is executed by a processor, it implements an optimization method for energy storage site selection and distribution network collaborative loss reduction based on photovoltaic-energy storage as described above.

[0175] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0176] Specific examples have been used to illustrate the principles and implementation methods of this invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this invention. Furthermore, those skilled in the art will recognize that, based on the ideas of this invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A method for optimizing energy storage site selection and distribution network loss reduction based on photovoltaic-energy storage, characterized in that, Includes the following steps: S1. By introducing voltage-current coupling coefficient and temperature correction factor, an improved line loss calculation model is constructed to quantify the nonlinear effects of power factor and the dynamic characteristics of line resistance with temperature. S2. Calculate the photovoltaic access line loss sensitivity factor of each node, identify the key nodes that are sensitive to line loss, and use them as the site selection for the energy storage system (ESS). S3. Based on the improved line loss calculation model and the location of the energy storage system ESS, an adaptive line loss minimization control strategy is established. Combined with the line loss-voltage joint optimization objective function, the expected value of line loss, variance and voltage stability are comprehensively weighed. S4. The differential dynamic programming method is used to optimize the ESS charge and discharge curve of the energy storage system. Combined with the high-dimensional alternating direction multiplier method for distributed solution, the collaborative loss reduction of multiple energy storage systems and distribution network is realized.

2. The method for optimizing energy storage site selection and distribution network loss reduction based on photovoltaic-energy storage according to claim 1, characterized in that, In S1, the improved line loss calculation model is as follows: Where, ΔP Loss_m1 R is the line loss between node m-1 and node m. m (T) represents the line resistance between node m-1 and node m at temperature T, U m P is the effective value of the voltage at node m. j Let Q be the active power load of node j. j For the reactive power load at node j, P pv Q represents the active power output of the photovoltaic system. pv For the reactive power output of the photovoltaic system, These represent the active power during charging, active power during discharging, reactive power during charging, and reactive power during discharging of the energy storage system at time t, respectively, ω. t For time-varying weighting coefficients, SOC ess For the energy storage system, f(SOC) represents the state of charge. ess ) is a function of the energy storage state of charge.

3. The method for optimizing energy storage site selection and distribution network loss reduction based on photovoltaic-energy storage according to claim 1, characterized in that, In step S2, the photovoltaic access line loss sensitivity factor of each node is calculated to identify key nodes sensitive to line loss, specifically including: After the distributed photovoltaic system is connected to the power distribution network, the active power and reactive power of the photovoltaic access node are corrected. Based on the total line resistance from the node with load to the root node, the sensitivity of power change to line loss is quantified to obtain the photovoltaic access line loss sensitivity factor. Select the node with the highest average value of photovoltaic access line loss sensitivity factor within 24 hours to deploy ESS.

4. The method for optimizing energy storage site selection and distribution network loss reduction based on photovoltaic-energy storage according to claim 3, characterized in that, S2 also includes introducing a line loss sensitivity factor for load fluctuations to address different load characteristics: Wherein, ΔP Loss ΔL represents the change in line loss. i The load change at node i, × i σ(L) is the weighting coefficient for node positions. i Let L be the standard deviation of the load at node i. i Let be the load power of node i.

5. The method for optimizing energy storage site selection and distribution network loss reduction based on photovoltaic-energy storage according to claim 1, characterized in that, In S3, the objective function of the adaptive line loss minimization control strategy is as follows: Where minJ is the objective function, t is the time series index, T is the total number of time intervals within the optimization period, and λ t μ is the weighting coefficient for the expected value of line loss at time t. t v is the weighting coefficient for the variance of the line loss at time t. t Let E{ΔP} be the weighting coefficient of the volatility sensitivity factor at time t. Loss,t Let} be the expected value of the line loss at time t, Var{ΔP} Loss,t Let VSF be the variance of the line loss at time t. t Let be the volatility sensitivity factor at time t.

6. The method for optimizing energy storage site selection and distribution network loss reduction based on photovoltaic-energy storage according to claim 4, characterized in that, The objective function for the joint optimization of line loss and voltage is: Where α is the line loss weighting coefficient, β is the mean square voltage deviation weighting coefficient, γ is the maximum voltage deviation weighting coefficient, and ΔP Loss,0 For baseline loss, U ref The reference voltage value is N, and the total number of nodes in the distribution network is N.

7. The method for optimizing energy storage site selection and distribution network loss reduction based on photovoltaic-energy storage according to claim 1, characterized in that, In step S4, the optimization of the ESS charge-discharge curve using a differential dynamic programming method includes: Define the value function V(x) T ) and terminal cost function φ(x T The control variables are updated iteratively using incremental dynamic programming, as shown in the following formula: in, This represents the update amount of the control variable in the (i+1)th iteration. Let be the second derivative of the Hamiltonian function with respect to the control variable. The first derivative of the Hamiltonian function with respect to the control variable is... This is the mixed second derivative of the Hamiltonian function with respect to the control and state variables. Update the state variables for the (i+1)th iteration.

8. The method for optimizing energy storage site selection and distribution network loss reduction based on photovoltaic-energy storage according to claim 1, characterized in that, In S4, the distributed solution using the high-dimensional alternating direction multiplier method includes: The global optimization problem is decomposed into multiple subproblems, and a distributed solution using the high-dimensional alternating direction multiplier method is employed. Where x represents the voltage adjustment and reactive power compensation at each node, z represents the charging and discharging power of the energy storage system, and y represents the Lagrange multiplier corresponding to the power balance constraint; collaborative optimization is achieved by augmenting the Lagrange function. for: Where f(x) is the line loss objective function related to grid voltage and reactive power, g(z) is the cost and loss function related to energy storage system charging and discharging, Ax is the power equivalent of grid node regulation, Bz is the power injection amount of ESS charging and discharging, c is the baseline value of system net power demand, and y T Let y be the dual variable and ρ be the penalty parameter.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the photovoltaic-energy storage-based energy storage location selection and distribution network collaborative loss reduction optimization method as described in any one of claims 1 to 8.

Images