Power distribution network source load storage coordination control method based on distributed power supply, electronic equipment and medium

Abstract

The invention discloses a power distribution network source load storage coordination control method based on a distributed power supply, electronic equipment and a medium. The method comprises the following steps: acquiring operation data of a power distribution network in real time; judging based on a preset switching condition and the operation data of the power distribution network, if the operation of the power distribution network does not meet the switching condition, controlling the power distribution network to execute a centralized regulation and control strategy, and obtaining an optimal global control strategy, the preset switching condition at least comprising communication interruption; based on the optimal global control strategy, controlling a plurality of sub-regions to execute a first regulation and control instruction, wherein the sub-regions comprise user distributed photovoltaic power generation; if the operation of the power distribution network meets the switching condition, controlling the power distribution network to execute a distributed regulation and control strategy, and obtaining an optimal local control strategy corresponding to each sub-region; and controlling the corresponding sub-region to execute the second regulation and control instruction based on the optimal local control strategy. The centralized regulation and control and the distributed regulation and control are switched through the preset switching condition, so that the response speed and the flexibility can be improved, and the intervention cost can be reduced.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of power distribution network of power system, and particularly relates to a distribution network source-load-storage coordinated control method based on distributed power sources, an electronic device and a medium. BACKGROUND

[0002] With the increasing proportion of renewable energy in the power grid system, especially the popularity of solar energy, the stability and flexibility of the power grid are facing more and more challenges. The volatility of solar energy makes the dispatching of the power grid more complex.

[0003] In the related art, the source-grid-load-storage collaborative control strategy usually relies on single centralized scheduling or distributed scheduling. Such scheduling method has problems of slow response speed, poor flexibility, high system intervention cost, etc., and is difficult to cope with load fluctuation and rapidly changing power demand. SUMMARY

[0004] The purpose of the present application is to switch between centralized scheduling and distributed scheduling through preset switching conditions, so as to realize a distribution network source-load-storage coordinated control method based on distributed power sources, which improves the response speed and flexibility and reduces the intervention cost.

[0005] In order to achieve the above-mentioned purpose, the present application provides a distribution network source-load-storage coordinated control method based on distributed power sources, comprising: acquiring real-time distribution network operation data, the distribution network operation data including power grid load data, electricity demand data of different regions and power change data of energy storage devices; based on preset distribution network control switching conditions and the distribution network operation data, if the distribution network operation does not meet the preset distribution network control switching conditions, control the distribution network to execute a centralized control strategy, and obtain an optimal global control strategy to realize load balancing, maximize renewable energy resource utilization and reduce cost, the centralized control strategy includes a first multi-objective optimization model based on constraint conditions, the preset distribution network control switching conditions at least include communication interruption; based on the optimal global control strategy, control a plurality of sub-regions to execute a first control instruction, the sub-regions include user distributed photovoltaic power generation; if the distribution network operation meets the preset distribution network control switching conditions, control the distribution network to execute a distributed control strategy, and obtain an optimal local control strategy corresponding to each sub-region to realize global load balancing; based on the optimal local control strategy, control the corresponding sub-region to execute a second control instruction.

[0006] In an optional embodiment, the preset distribution network control switching conditions further include that the power grid load fluctuation exceeds a load fluctuation threshold and the power grid load exceeds a load threshold, and the communication interruption includes that the communication delay exceeds a delay threshold and the network packet loss rate exceeds a packet loss rate threshold; wherein the expression that the power grid load fluctuation exceeds the load fluctuation threshold is: |Pload (t) - P load (t - 1) > ΔP threshold ; wherein, P load (t) and P load (t - 1) are the power grid load at time t and t - 1 respectively, ΔP threshold is a preset load fluctuation threshold; the expression that the communication delay exceeds a delay threshold is: Delay(t) > ΔDelay threshold ; wherein, Delay(t) is the current communication delay, ΔDelay threshold is a preset maximum communication delay threshold; the expression that the network packet loss rate exceeds a packet loss rate threshold is: ; wherein, P lost is the number of lost packets, P total is the total number, ΔP rate is the packet loss rate threshold; the expression that the power grid load exceeds a load threshold is: P load (t) > a · P max ; wherein, P load (t) is the current power grid load, P max is the maximum load of the power grid, and a is a load ratio.

[0007] In an optional embodiment, the expression of the first multi-objective optimization model is:

[0008] ; wherein, C gen,i is the resource cost of the i-th region, C stor,i is the charging and discharging cost of the energy storage system, C solar,i is the cost related to the power generated by the solar panel of the user, L bal is the cost of load balancing so that the power grid load and the generated power are balanced, R gen is the utilization rate of renewable energy, and λ1 and λ2 are adjustment parameters; the expression of the constraint condition is: ; wherein, P load,i is the load demand of the i-th region, P gen,i is the generated power, P stor,i is the power of the energy storage device, and P solar,i is the output power of the distributed photovoltaic power generation; the expression of the safety constraint of the energy storage device is: ; wherein, SOC i (t) is the battery state of the energy storage device, SOC min is the minimum energy storage of the energy storage device, and SOC max is the maximum energy storage of the energy storage device.

[0009] In an alternative embodiment, the distributed regulation control strategy comprises a second multi-objective optimization model based on the constraint condition, and the expression of the second multi-objective optimization model is min f i (x i )=C stor,i +C solar,i ; in the formula, C stor,i is the cost of the energy storage device, C solar,i is the related cost of the distributed photovoltaic power generation, and x i is the scheduling decision variable of the ith sub-region.

[0010] In an alternative embodiment, if the power distribution network operation meets the preset power distribution network control switching condition, the power distribution network is controlled to execute the distributed regulation control strategy, and the optimal local control strategy corresponding to each sub-region is obtained, specifically including: initializing the decision variable, the dual variable and the global decision variable of each sub-region; obtaining the local data of the sub-region in real time; calculating the gradient based on the local data and the second multi-objective optimization model:

[0011] in the formula, a is the learning rate, is the gradient of the second multi-objective optimization model of the ith sub-region, z k is the global scheduling target, is the dual variable of the ith sub-region, is the decision variable of the ith sub-region in the kth iteration, and p is the penalty factor; the sub-region is preliminarily scheduled based on the gradient; the scheduling result is obtained based on the preliminary scheduling; the scheduling result is shared with other sub-regions, and the optimal local control strategy is obtained by solving based on the local data and the second multi-objective optimization model.

[0012] In an alternative embodiment, any of the power distribution network source-load-storage coordination control methods based on distributed power sources further comprises: in the case of controlling the power distribution network to execute the centralized regulation control strategy, constructing a third multi-objective optimization model based on the power distribution network operation data: in the formula, C gen,i is the power generation cost of the ith sub-region, C stor,i is the charging and discharging cost of the energy storage device of the ith sub-region, C solar,i is the related cost of the distributed photovoltaic power generation of the ith region, L bal is the load balance constraint term, R gen is the comprehensive utilization rate of renewable energy, and 1-R genThe system is encouraged to maximize the utilization of renewable energy, with λ1 and λ2 being weighting coefficients. The third multi-objective optimization model is solved to obtain the optimal distributed photovoltaic power generation call amount. If the optimal distributed photovoltaic power generation call amount is greater than zero, then the distributed photovoltaic power generation feedback is invoked.

[0013] In an optional implementation, any of the distributed power generation-load-storage coordinated control methods for distribution networks further includes: implementing a distributed regulation and control strategy in the control distribution network; constructing a dynamic electricity price based on the overall network load forecast and the distribution network operation data: p(t) = αL(t) + βR(t) + γ·SOC(t). p(t) is the electricity price at time t, L(t) is the overall network load forecast at time t, R(t) is the renewable energy generation at time t, SOC(t) is the charging status of electric vehicles or energy storage systems, α is the load weighting coefficient, β is the renewable energy weighting coefficient, and γ is the energy storage status weighting coefficient; constructing a grid scenario based on the dynamic electricity price signal, the grid scenario including peak load periods, periods of low single load with sufficient sunlight, periods of high output from other renewable energy sources, and periods of low generation from other renewable energy sources; increasing or decreasing the electricity price based on the grid scenario to incentivize distributed photovoltaic power generation users to feed electricity back to the grid.

[0014] In one optional implementation, the distribution network source-load-storage coordinated control method based on distributed generation further includes: constructing a reward rule: R reward (t)=δ·P feedback (t)·η(t)·T(t); where R reward P(t) represents the user's reward at time t. feedback η(t) represents the electrical energy fed back to the grid by the user, η(t) represents the grid load factor, T(t) represents the user's feedback duration, and δ represents the reward coefficient, which is the basic amount for adjusting the reward. Based on the current grid status and the dynamic electricity price, a real-time electricity price and the reward rules are calculated and published, so that users can adjust their behavior of feeding back electricity to the grid according to the received real-time electricity price and reward rules. The deviation between the user's behavior of feeding back electricity to the grid and the expected target is analyzed; if the user's discharge is insufficient, the electricity price and reward are increased; if the user's discharge is sufficient, the electricity price and reward are decreased.

[0015] The present invention also proposes an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform any of the power generation, load and storage coordination control methods for distributed power grids of distributed generation.

[0016] The application further provides a medium storing a computer program, which, when executed by a processor, implements the power distribution network source-load-storage coordination control method based on distributed power sources.

[0017] The application has the advantages that: when the power distribution network operation meets the switching condition, the distributed regulation and control strategy is switched in time, thereby ensuring the continuous and stable operation of the power grid under special conditions. Through distributed regulation and control, each sub-region can formulate an optimal local control strategy according to its own operation state, can quickly respond to local power supply and demand changes, and realizes the optimal allocation and flexible scheduling of local resources. Through information interaction and collaborative cooperation between sub-regions, the overall load balance is finally achieved, the local optimization gradually realizes the overall load balance, the risk resistance and self-healing capabilities of the power grid are greatly enhanced, the ability of the power grid to cope with emergencies and complex operating environments is improved, the risk resistance and self-healing capabilities of the power grid are enhanced, and the power grid can still maintain basic power supply functions and operation stability under extreme conditions. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 A flowchart of the power distribution network source-load-storage coordination control method based on distributed power sources provided by the embodiment of the application is provided.

[0019] Figure 2 A flowchart of the power distribution network source-load-storage coordination control method based on distributed power sources provided by another embodiment of the application is provided.

[0020] Figure 3 A real-time electricity price and incentive mechanism flowchart of the power distribution network source-load-storage coordination control method based on distributed power sources provided by the embodiment of the application is provided. DETAILED DESCRIPTION

[0021] With the popularization of solar panels, the difficulty of power grid load regulation increases, especially when the light is strong, the power grid may face the situation of power over-supply or overload.

[0022] In order to cope with the above challenges, and considering that more and more users install solar panels at present, the solar energy generated by user households is taken as one of the distributed resources of the power grid regulation system. However, in the existing system, the enthusiasm of users to participate in the feedback of distributed photovoltaic power generation to the power grid is low, and there is a lack of sufficient incentive and dynamic compensation mechanism. Therefore, how to design a dynamic incentive mechanism to encourage users to autonomously regulate the charging and discharging behavior of their solar power generation according to the demand of the power grid is a problem to be solved in the current technology.

[0023] The application will be further described in detail through the accompanying drawings and specific embodiments.

[0024] As Figure 1 and Figure 2As shown, according to the embodiment of the application, in one aspect, a distributed power supply based power distribution network source-load-storage coordination control method is provided, comprising the following steps:

[0025] Step S101: Real-time acquisition of power distribution network operation data, the power distribution network operation data including power grid load data, power demand data of different regions and power change data of energy storage devices.

[0026] Step S103: Based on the preset power distribution network control switching condition and the power distribution network operation data, if the power distribution network operation does not meet the preset power distribution network control switching condition, the power distribution network is controlled to execute a centralized regulation and control strategy, and an optimal global control strategy is obtained, so as to realize load balancing, maximize renewable energy resource utilization rate and reduce cost, the centralized regulation and control strategy including a first multi-objective optimization model based on constraint conditions, and the preset power distribution network control switching condition at least including communication interruption.

[0027] Step S105: Based on the optimal global control strategy, a first regulation and control instruction is executed in multiple sub-regions, and the sub-regions are interacted and cooperated through information interaction, and the sub-regions include user distributed photovoltaic power generation.

[0028] Step S107: If the power distribution network operation meets the preset power distribution network control switching condition, the power distribution network is controlled to execute a distributed regulation and control strategy, and an optimal local control strategy corresponding to each sub-region is obtained, so as to realize global load balancing.

[0029] Step S109: Based on the optimal local control strategy, a second regulation and control instruction is executed in the corresponding sub-region.

[0030] The embodiment of the application provides a two-layer architecture, including a centralized decision layer and a regional control layer.

[0031] The core target of the centralized control layer is to realize all-round and fine optimal allocation of global network resources, so as to ensure that the power grid can maintain load balancing and efficiently utilize renewable energy in a complex and changeable operation environment, thereby realizing orderly and coordinated feedback of user distributed photovoltaic power generation.

[0032] The central scheduling platform in the centralized control mode collects data in all corners of the global network in real time and comprehensively through a high-speed and stable communication network. These data include real-time change of power grid load, power demand of different regions, power change of energy storage devices and other key information. Through deep analysis and mining of these massive data, the scheduling platform can accurately grasp the operation situation of the power grid, and provide a solid data foundation for subsequent optimization scheduling decision. The centralized control mode adopts a multi-objective optimization algorithm, which can scientifically and reasonably weigh and optimize multiple interrelated and possibly conflicting targets, and is suitable for new energy power systems.

[0033] The resource scheduling of new energy power systems needs to consider different types of power generation methods, such as wind power, distributed photovoltaic power generation, traditional thermal power, etc., whose generation costs are different. The centralized control strategy needs to consider the cost factors of various power generation methods, optimize the generation combination, and reduce the overall generation cost as much as possible. The use of renewable energy emphasizes the maximum use of clean energy such as wind energy and distributed photovoltaic power, improves the proportion of renewable energy in power supply, and promotes the green transformation of energy structure. The centralized control strategy will adjust the generation plan according to the real-time generation capacity and prediction data of renewable energy, prioritize the grid-connected generation of renewable energy, and reduce the dependence on traditional fossil energy. The charging and discharging target of energy storage equipment is mainly to realize the peak load shifting of power and improve the stability and energy utilization efficiency of the power grid by reasonably controlling the charging and discharging process of energy storage equipment. In order to balance the relationship between these different goals, the first multi-objective optimization model that can realize multi-objective optimization algorithm is adopted.

[0034] Solving the first multi-objective optimization model will obtain the optimal global control strategy, which can be refined into the first control instruction for each sub-region, ensuring the effective implementation of the control strategy. Including user distributed photovoltaic power generation in the sub-region control range makes the distributed power better integrated into the overall operation system of the power grid, and fully utilizes the potential of user distributed photovoltaic power generation. Through accurate control, the coordination between distributed power and large power grid is realized, the consumption capacity of distributed power is improved, the adaptability and inclusiveness of the power grid to distributed energy are enhanced, and the flexibility and reliability of distribution network operation are also improved.

[0035] To ensure the stable operation of the system, constraint conditions are constructed for the first multi-objective optimization model.

[0036] The communication link of the power grid system is the key part of centralized control. When communication fails, centralized control cannot obtain real-time data or issue instructions, and the system needs to automatically switch to distributed control mode.

[0037] Under the distributed control model of regional control layer, the control system in each region shows high autonomy and flexibility, and makes accurate and timely optimization decisions according to local real-time data.

[0038] When the power distribution network operation meets the switching condition (such as communication interruption), the distributed regulation and control strategy is switched in time to ensure the continuous and stable operation of the power grid under special conditions. Through distributed regulation and control, each sub-area can develop the optimal local control strategy according to its own operation state, giving the sub-area a high degree of autonomy, enabling it to quickly respond to local power supply and demand changes, and achieving optimal allocation and flexible scheduling of local resources. Through information exchange and coordination between sub-areas, the overall load balance is ultimately achieved, gradually realizing load balance from local optimization to overall optimization, greatly enhancing the anti-risk and self-healing capabilities of the power grid, improving the ability of the power grid to respond to sudden conditions and complex operating environments, enhancing the anti-risk and self-healing capabilities of the power grid, and even under adverse conditions, the power grid can still maintain basic power supply functions and operational stability.

[0039] The regional control layer plays a key role in local optimization in the operation of the power grid system. When the power grid encounters large load fluctuations or communication interruption occurs, the system switches to a distributed control mode.

[0040] Further, the preset power distribution network control switching condition further includes that the power grid load fluctuation exceeds a load fluctuation threshold and the power grid load exceeds a load threshold, and the communication interruption includes that the communication delay exceeds a delay threshold and the network packet loss rate exceeds a packet loss rate threshold.

[0041] Among them, the change amplitude of the power grid load is set as the basis for switching. If the power grid load changes by more than a preset threshold within a certain time (such as 1 minute or 5 minutes), the system will trigger a switch to distributed control. For example, when the load change exceeds 10% (or adjusted according to specific needs), the system will trigger a switch.

[0042] The expression that the power grid load fluctuation exceeds the load fluctuation threshold is:

[0043] |P load (t)-P load (t-1)|>ΔP threshold ;

[0044] In the formula, P load (t) and P load (t-1) are the power grid loads at time t and t-1, respectively, and ΔP threshold is a preset load fluctuation threshold.

[0045] The control mode can be switched by monitoring the network delay. When the communication delay exceeds the preset threshold, it means that the system may face communication problems, at which time the system will automatically switch to distributed control. The expression that the communication delay exceeds the delay threshold is:

[0046] Delay(t) > ΔDelay threshold ;

[0047] In the formula, Delay(t) is the communication delay at the current time, and ΔDelay threshold This is the preset maximum communication latency threshold.

[0048] Alternatively, the system can monitor the packet loss rate to determine if communication is severely disrupted. If the packet loss rate exceeds a certain threshold, the system will consider the communication interruption to be serious and switch to distributed control mode. The expression for the network packet loss rate exceeding the packet loss rate threshold is:

[0049]

[0050] In the formula, P lost P represents the number of packets lost. total For the total number of cases, ΔP rate This is the packet loss rate threshold.

[0051] When the power grid load approaches or exceeds its design capacity, centralized control may not be able to adjust the load across all areas in real time. In this situation, distributed control can respond quickly through local scheduling. The expression for when the power grid load exceeds the load threshold is:

[0052] P load (t)>α·P max ;

[0053] In the formula, P load (t) represents the current grid load, P max The maximum load of the power grid is α, which is the load ratio, for example, set to 0.8.

[0054] With clear mathematical expressions, it can be accurately and quantitatively judged whether the power grid is in a state of excessive load, excessive load fluctuation, poor communication condition (excessive delay, excessive packet loss rate), and whether the current power grid load exceeds the reasonable range can be quickly known to avoid misjudgment of the power grid state caused by manual or fuzzy judgment, and to provide accurate basis for subsequent control strategy switching. When the power grid meets any of the above switching conditions, it can be switched from centralized control to distributed control in time. In the case of good communication and stable load in centralized control, if the communication delay is too high or the packet loss rate is too high, the distributed control can rely on local autonomous decision to ensure that the power grid can still be effectively dispatched; when the load is too high or the load fluctuation is too large, the distributed control can quickly respond to local load changes to maintain the stability of the power grid and improve the ability of the power grid to cope with complex working conditions. The explicit preset power distribution grid control switching condition can match the most suitable control strategy for the power grid in different operating states, reduce the power distribution grid operation problems caused by mismatched control strategies, such as load imbalance and renewable energy consumption difficulty, and thus improve the stability and reliability of the power distribution grid operation as a whole, ensuring the continuous and stable supply of electricity.

[0055] Further, the expression of the first multi-objective optimization model is:

[0056]

[0057] In the formula, C gen,i is the resource cost of the i-th region, such as generation cost, energy storage cost, etc., C stor,i is the charging and discharging cost of the energy storage system, C solar,i is the related cost of the user's solar panel generated power, such as maintenance, equipment depreciation, etc., L bal is the cost of load balancing to balance the power grid load and generated power, R gen is the utilization rate of renewable energy, maximizing the utilization of solar and wind power generation, and λ1 and λ2 are adjustment parameters to adjust the priority of different objectives, which are dynamically adjusted according to the real-time operating conditions of the power grid, energy policy orientation, and economic cost and other factors.

[0058] To ensure stable operation of the system, the load of the power grid system should be balanced with the power generation, energy storage device discharge and distributed photovoltaic power generation.

[0059] The constraint conditions include load balancing constraint and energy storage device safety constraint, and the expression of the load balancing constraint is:

[0060]

[0061] In the formula, P load,i is the load demand of the i-th region, P gen,i is the power generation, and P stor,iP solar,i is the output power of the distributed photovoltaic power generation.

[0062] Each energy storage device should be kept within a safe range, and the expression of the energy storage device safety constraint is:

[0063]

[0064] SOC i (t) is the battery state of the energy storage device, SOC min is the minimum energy storage of the energy storage device, SOC max is the maximum energy storage of the energy storage device.

[0065] The control system of the sub-area first constructs a corresponding local optimization objective function according to the locally collected data, and this objective function reflects the goal that the sub-area hopes to achieve through scheduling optimization under the current operating state.

[0066] Further, the expression of the second multi-objective optimization model is:

[0067] minf i (x i )=C stor,i +C solar,i ;

[0068] C stor,i is the cost of the energy storage device, C solar,i is the related cost of the distributed photovoltaic power generation, x i is the scheduling decision variable of the i-th sub-area, such as the discharge amount of the energy storage device, the solar energy feedback amount, etc.

[0069] In addition, the sub-area coordinated optimization also needs to meet the load balance constraint and the SOC constraint.

[0070]

[0071] P load,i is the load demand of the i-th area; P gen,i is the power generation; P stor,i is the power of the energy storage device; P solar,i is the output power of the distributed photovoltaic power generation.

[0072] Each energy storage device should be kept within a safe range, and the mathematical model is as follows:

[0073]

[0074] SOC i (t) is the battery state of the energy storage device; SOC min is the energy storage device.

[0075] Then, the distributed regulation and control strategy calculates the gradient of the second multi-objective function based on the local data.

[0076] Under the distributed control model of the regional control layer, the control system in each region exhibits high autonomy and flexibility, and makes accurate and timely optimization decisions according to real-time local data. In local control, the distributed gradient descent algorithm (DGD) is the core technical support for realizing efficient optimization scheduling. The scheduling task of each sub-region is optimized under the framework of the DGD algorithm.

[0077] Further, based on step S107, if the power distribution network operation meets the preset power distribution network control switching condition, the power distribution network is controlled to execute the distributed regulation and control strategy, and the optimal local control strategy corresponding to each sub-region is obtained, specifically including the following steps:

[0078] Step S1071: initialize the decision variable, dual variable and global decision variable of each sub-region;

[0079] Step S1073: real-time acquisition of local data of the sub-region.

[0080] Step S1075: based on the local data and the second multi-objective optimization model, the gradient is calculated:

[0081]

[0082] In the formula, α is the learning rate, is the gradient of the second multi-objective optimization model of the i th sub-region, z k is the global scheduling target, is the dual variable of the i th sub-region, is the decision variable of the i th sub-region in the k th iteration, and ρ is the penalty factor.

[0083] Step S1077: preliminary scheduling of the sub-region based on the gradient.

[0084] Step S1078: obtaining the scheduling result based on the preliminary scheduling.

[0085] Step S1079: sharing the scheduling result with other sub-regions, solving based on the local data and the second multi-objective optimization model, and obtaining the optimal local control strategy.

[0086] In this embodiment, the gradient is a vector, the direction of which represents the direction in which the objective function rises fastest at the current power, and the opposite direction is the direction in which the objective function descends fastest. By calculating the gradient, the distributed regulation and control strategy of the sub-region can understand the gap and direction between the current scheduling strategy and the optimal strategy, and the expression of the gradient is as follows:

[0087] where a is the learning rate, determining the step size of each iteration; is the gradient of the ith sub-region objective function, indicating the optimization direction of the current decision; k is the global scheduling objective, such as the optimal allocation of the entire network load; is the dual variable of the ith sub-region, ensuring that the local decision is consistent with the global objective; is the decision variable of the ith sub-region in the kth iteration, such as the discharge amount of energy storage devices or the feedback amount of distributed photovoltaic power; p is the penalty factor, controlling the coordination degree between local decision and global decision.

[0088] First, the distributed regulation control strategy initializes the decision variables of each sub-region such as the initial power feedback, dual variable and global decision variable z 0 which is the load balancing objective of the entire network. Then each sub-region i calculates the gradient of the objective function according to the local data and updates the decision variable using the gradient descent method In each iteration, the sub-region passes its local optimization result to other sub-regions through an information exchange mechanism. These information help the entire network to coordinate and adjust the scheduling strategy of each sub-region, ensuring the load balancing of the entire network. The dual variable will be adjusted according to the current global scheduling situation, helping the local decision to be consistent with the global optimization objective. The global optimization objective z k Through information exchange and the role of dual variable, the local decision of each sub-region is gradually adjusted, making the system overall coordinated. The global scheduling result will be optimized through multiple iterations, finally ensuring the optimal scheduling of the entire network load. With the iteration, the local decision of all sub-regions gradually approaches the optimal solution of the entire network load, thus realizing the load balancing and resource optimization of the power grid. After completing the preliminary scheduling, each sub-region control system also needs to share its scheduling result with other sub-regions in time. Through this information interaction, even without global control, each sub-region can work together to achieve the optimal load scheduling of the entire network.

[0089] Distributed photovoltaic power generation users can feed back power to the grid when their power generation is greater than their own consumption. Distributed photovoltaic power generation systems need to work closely with wind power, hydropower and other renewable energy systems in grid dispatching to form a synergistic effect. Since various types of renewable energy all have volatility, their generation characteristics need to be considered comprehensively for optimized dispatching. For example, during sunny days, the output of distributed photovoltaic power generation is large, and the generation plan of wind power and hydropower can be appropriately reduced; when it is night or the light is insufficient, the output of wind power and hydropower can be increased to make up for the deficiency of distributed photovoltaic power generation. Through overall coordination of multi-level dispatching control, the advantages of various types of new energy can be fully utilized, the overall output curve can be smoothed, and the acceptance capacity and operation economy of the grid for renewable energy can be improved.

[0090] In order to accurately evaluate and predict the output of distributed photovoltaic power generation, the distributed photovoltaic power generation system of each user needs to be modeled. The solar power generation mainly depends on the conversion efficiency of the solar panel and the real-time light intensity. The modeling expression of the user's distributed photovoltaic power generation is as follows:

[0091] P solar (t)=η·A·G(t)·(1-α);

[0092] Wherein, P solar,i (t) is the power generated by the solar panel of the i-th user; G(t) is the light intensity; η is the efficiency of the solar panel; and α is the system comprehensive loss coefficient.

[0093] When the grid load is low, the power of distributed photovoltaic power generation will first be used to meet the local demand (such as household electricity), and the remaining power can be stored or fed back to the grid.

[0094] When the grid load is high, the system encourages users to feed back the stored power or the remaining solar power generation to the grid to relieve the load pressure of the grid.

[0095] In the centralized control mode, the central platform balances the grid load and maximizes the utilization of renewable energy (including distributed photovoltaic power, wind power, hydropower, etc.) through a global optimization algorithm, and formulates a dispatching strategy for the whole grid.

[0096] In this mode, the system will collect real-time information such as grid load data, regional solar power generation, wind power and hydropower output, etc., and through building and solving the following grid dispatching model, it is determined whether to call the feedback of the user's distributed photovoltaic power generation.

[0097] Further, the source-load-storage coordination control method based on the distributed power supply further includes the following steps:

[0098] Step S201: In the case of controlling the distribution network to execute a centralized control strategy, a third multi-objective optimization model is constructed based on the operation data of the distribution network:

[0099]

[0100] C gen,i C stor,i C solar,i L bal R gen 1-R gen 1-R gen 1-R

[0101] Step S203: solving the third multi-objective optimization model to obtain the optimal distributed photovoltaic power generation calling amount.

[0102] Step S205: if the optimal distributed photovoltaic power generation calling amount is greater than zero, calling the distributed photovoltaic power generation feedback.

[0103] In this embodiment, by solving the third multi-objective optimization model, the centralized control layer can obtain the optimal distributed photovoltaic power generation calling amount, wind power and water power output plan, and energy storage device charging and discharging strategy, etc., to ensure the economy and stability of the whole network operation.

[0104] The optimal distributed photovoltaic power generation calling amount: it is clear that how much distributed photovoltaic power generation amount needs to be called from the user to participate in the power supply and balance of the power grid.

[0105] The wind power and water power output plan: it determines the output arrangement of renewable energy sources such as wind power and water power, that is, when and how much power to supply to the power grid.

[0106] The energy storage device charging and discharging strategy: it obtains the charging and discharging plan of the energy storage device, including when to charge, when to discharge, and the power size of charging and discharging, etc., so as to utilize the energy storage device to adjust the power supply and demand of the power grid, and to ensure the stability and economy of the distribution network operation.

[0107] When the optimal distributed photovoltaic power generation calling quantity obtained by solving is greater than 0, it indicates that the model considers that the distributed photovoltaic power generation needs to be supplemented to meet the power grid demand (such as balancing the load and improving the renewable energy utilization rate) after calculation, and the photovoltaic power generation feedback of the user is called at this time; if the calling quantity is 0, it indicates that the existing wind power, hydropower output and energy storage adjustment can meet the requirements of the distribution network operation, and there is no need to call the photovoltaic power generation feedback.

[0108] If the solving result shows that the wind power output is lower than expected or the energy storage device is in a discharge saturation / needs to be charged preferentially, it will cause the gap of available power of the power grid to increase, and the model is more likely to obtain a result greater than 0 when calculating the optimal distributed photovoltaic power generation calling quantity, thereby triggering the calling of the photovoltaic power generation feedback.

[0109] If the solving result shows that the wind power output is sufficient and the energy storage device can supplement power through discharge, the balance pressure of the power supply and demand of the power grid is small, and the optimal distributed photovoltaic power generation calling quantity calculated by the model is more likely to be 0, so that the photovoltaic power generation feedback does not need to be called.

[0110] When the power grid encounters a large load fluctuation or a local communication fault, the system will switch to a distributed control mode. At this time, the control system of each region will make independent optimization decisions based on the locally collected data to maintain the stable operation of the local power grid. For distributed photovoltaic power generation, the regional control system will focus on local light conditions, distributed photovoltaic power generation, user electricity demand and energy storage device state information.

[0111] The local dispatching decision will specifically consider the feedbackable amount of distributed photovoltaic power generation in each region, the maximum charge and discharge power and the remaining capacity of the energy storage device and other constraint conditions, so as to determine whether the local distributed photovoltaic power is preferentially stored or fed back to other loads in the region or the upper level.

[0112] In order to fully mobilize the user's distributed photovoltaic power generation to participate in the distributed photovoltaic power generation feedback, the distributed regulation and control strategy integrates dynamic electricity price and incentive mechanism.

[0113] Among them, the distributed regulation and control strategy is based on the whole network load prediction and real-time operation data, and guides the photovoltaic power generation behavior of the user through the dynamic electricity price signal.

[0114] Further, the source-load-storage coordinated control method of the distribution network based on the distributed power supply further includes the following steps:

[0115] Step S301: control the distribution network to execute the distributed regulation and control strategy, and construct a dynamic electricity price based on the whole network load prediction and the operation data of the distribution network:

[0116] p(t) = aL(t) + bR(t) + gSOC(t).

[0117] p(t) is the electricity price at time t; L(t) is the total network load prediction at time t; R(t) is the power generation of renewable energy at time t; SOC(t) is the state of charge (SOC) of the electric vehicle or energy storage system, which determines the charging or discharging capacity of the battery; a is a load weight coefficient (positive value), which reflects the positive effect of load on electricity price, and the higher the load, the more obvious the electricity price increase; β is a renewable energy weight coefficient (negative value), which reflects the negative effect of renewable energy output on electricity price, that is, the more renewable energy, the more the electricity price decreases; γ is an energy storage state weight coefficient, the sign and size of which can be set according to system requirements. For example, when the system needs more peak shaving resources, if the average SOC is high, it can be set to a negative value to encourage users to charge energy storage by reducing the electricity price; if discharging support is needed, it can be set to a positive value to encourage users to discharge by increasing the electricity price.

[0118] Step S303: Constructing a power grid scenario based on the dynamic electricity price signal, the power grid scenario including a load peak period, a sufficient light single load valley period, a period with high output of other renewable energy, and a period with low power generation of other renewable energy.

[0119] Step S305: Increasing or decreasing the electricity price based on the power grid scenario to encourage distributed photovoltaic power generation users to feedback power to the power grid.

[0120] In this embodiment, during the power grid load peak period, the real-time grid-connected electricity price is significantly increased and a special subsidy is added to encourage users to send the stored solar energy or real-time excess electricity back to the power grid to obtain high income and effectively help to reduce peak load.

[0121] During the period of sufficient light but low power grid load, the system reduces the grid-connected electricity price to guide users to preferentially use photovoltaic power generation for local energy storage and reduce power feeding. At the same time, users who store energy during this period are given a subsidy to encourage rational use of resources and avoid power grid voltage fluctuations and resource waste.

[0122] When the output of other renewable energy in the power grid system is high, the electricity price should be appropriately reduced to encourage users to charge or store electric energy.

[0123] When the power generation of other renewable energy is low, the electricity price should be increased to encourage users to reduce electric energy consumption or feedback electric energy to the power grid.

[0124] The incentive mechanism aims to guide users to feedback power to the power grid during the peak load period of the power grid through dynamic rewards, and adjusts the rewards according to the participation time, discharge amount and other factors of the user, to ensure the load balance and maximize the resource utilization of the system. The reward obtained by the user when participating in the power feedback is related to the following three factors: (1) discharge amount: the more power the user feedbacks, the higher the reward; (2) discharge time: users who provide power feedback during the peak load period of the power grid should obtain higher rewards; (3) participation duration: the longer the user participates in the power feedback, the reward can be appropriately increased.

[0125] Further, in combination with Figure 3 As shown in the figure, the source-load-storage coordinated control method of the distribution network based on the distributed power supply further includes the following steps:

[0126] Step S307: constructing a reward rule:

[0127] R reward (t)=δ·P feedback (t)·η(t)·T(t);

[0128] In the formula, R reward (t) is the reward of the user at time t, P feedback (t) is the amount of power feedback to the power grid by the user, η(t) is the load coefficient of the power grid, the higher the load, the higher the reward coefficient; T(t) is the participation duration of the user, δ is the reward coefficient, and is the basic quantity of adjusting the reward.

[0129] Step S309: calculating and publishing the real-time electricity price and the reward rule according to the current state of the power grid and the dynamic electricity price, so that the user adjusts the behavior of feedback power to the power grid according to the received real-time electricity price and the reward rule.

[0130] Step S311: analyzing the deviation of the behavior of the user feedback power to the power grid from the expected target.

[0131] Step S313: if the discharge amount of the user is insufficient, the electricity price and the reward are increased.

[0132] Step S315: if the discharge amount of the user is sufficient, the electricity price and the reward are reduced.

[0133] In this embodiment, the dynamic electricity price and the incentive mechanism complement each other and work together to form an economic lever system that guides user behavior and optimizes the operation of the distribution network.

[0134] During the peak load period of the power grid, the real-time electricity price p(t) is high at this time. At the same time, the load coefficient η(t) in the incentive mechanism will increase with L(t), resulting in that the reward R reward(t) significantly increases. The double stimulation of high electricity price and high reward will greatly mobilize the enthusiasm of users to participate in electricity feedback, attract more users to feed back the stored electricity or excess solar power to the grid, and jointly alleviate the peak pressure.

[0135] During the low load period of the power grid: the real-time electricity price p(t) is low, and η(t) in the incentive mechanism is also small, so the reward obtained by the user for discharging is correspondingly reduced. In this case, the electricity price signal and the incentive signal jointly guide the user to reduce discharging and instead charge or store energy to adapt to the supply and demand state of the power grid at this time. The entire system has strong adaptability. The power grid operation system automatically adjusts the coefficients a, β, γ in the electricity price model and the functional relationship between the coefficients δ and η(t) in the incentive model by monitoring the data of the power grid load L(t), the renewable energy output R(t), the user energy storage state SOC(t), and the user participation in real time.

[0136] The strategy of the electricity price and the incentive mechanism needs to develop corresponding adjustment strategies for different distribution network operation scenarios.

[0137] During the peak load period, the weight a is increased to make p(t) significantly increase, increase the electricity cost, and suppress unnecessary demand; the base reward coefficient δ and the load coefficient η(t) are increased to make R reward (t) significantly increases, and the user is strongly encouraged to feed back electricity; through high price difference and high reward, enough electricity feedback is attracted to balance the peak load and avoid power cut.

[0138] During the period of strong light and sufficient solar power generation, the weight β is reduced (more negative) to make p(t) decrease and encourage users to use more electricity; at this time, if the load of the power grid is not high, the discharging reward can be reduced or even suspended to guide users to store excess solar electricity in local energy storage devices instead of directly feeding it into the grid to cause pressure on the grid; at the same time, the utilization rate of renewable energy is maximized, and the renewable energy is converted into energy storage that can be used in subsequent periods, improving the utilization rate of renewable energy and reserving peak shaving resources for the power grid.

[0139] During the period of sudden drop of renewable energy output and insufficient power grid reserve, the weights a and β are increased (absolute value decreases) to make p(t) rapidly increase; the base reward coefficient δ and the load coefficient η(t) are urgently increased to start the emergency reward mechanism and encourage all users with discharging capacity to immediately feed back electricity to make up for the power supply gap. The goal is to quickly mobilize user-side resources to ensure the safe and stable operation of the power grid.

[0140] The implementation of dynamic electricity price and incentive mechanism is a closed-loop control system, and the core is to continuously optimize the control strategy according to the feedback of user behavior.

[0141] The power grid operation system calculates and publishes real-time electricity price p(t) and incentive rules according to the current power grid state; the user adjusts his own behavior according to the received electricity price and incentive signal, such as deciding when to use electricity, when to charge, when to discharge feedback, etc. For example, a household user may choose to put the electricity in the energy storage battery on the network at this time when he sees that the electricity price and the reward are both high during the peak period; the system will collect user behavior data in real time, such as the electricity consumption, the charging amount, the discharge feedback amount of each period, and the influence of these behaviors on the power grid state; then, the system analyzes the deviation of the user behavior from the expected target. If the user's discharge amount is insufficient during the load peak period and fails to effectively smooth the load, the electricity price and the reward of this period need to be further increased; if the solar energy is sufficient and the user's energy storage is insufficient, leading to light abandonment, the electricity price of this period needs to be further reduced, or an energy storage subsidy is provided. Through this feedback adjustment, the parameters such as alpha, beta, gamma and delta are continuously optimized, so that the control strategy is more in line with the user's behavior habits and more effectively achieves the optimization goal of the power grid.

[0142] In another aspect, the present application provides an electronic device, characterized in that it comprises: at least one processor; a memory in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform any one of the distributed power source-based distribution network source-load-storage coordination control methods.

[0143] In another aspect, the present application provides a medium, which is a computer storage medium, and stores a computer program, and the computer program is executed by a processor to implement any one of the distributed power source-based distribution network source-load-storage coordination control methods.

[0144] The computer storage medium can be referred to as a medium simply. Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware. The program can be stored in a non-volatile computer readable storage medium, and when the program is executed, the processes of the above-mentioned embodiments can be included. Any reference to memory, storage, database or other medium used in each embodiment provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM). Each embodiment in the specification is described in a progressive manner, and the same or similar parts of each embodiment can be referred to each other. Each embodiment focuses on the difference from other embodiments. Especially, for the device, equipment, non-volatile computer storage medium embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant part can be referred to the part of the method embodiment.

[0145] The above embodiments are only examples for clearly illustrating, and not limiting the embodiments. Based on the above description, other different forms of changes or variations can be made by those skilled in the art. Here, all the embodiments cannot be exhausted, and the obvious changes or variations still fall within the protection scope of the present application.

Claims

1. A method for coordinated control of power generation, load, and energy storage in a distribution network based on distributed generation, characterized in that, include: Real-time acquisition of power distribution network operation data, including power grid load data, electricity demand data in different regions, and power change data of energy storage devices; Based on the preset distribution network control switching conditions and the distribution network operation data, if the distribution network operation does not meet the preset distribution network control switching conditions, the distribution network is controlled to execute a centralized control strategy and obtain the optimal global control strategy to achieve load balance, maximize the utilization rate of renewable energy resources and reduce costs. The centralized control strategy includes a first multi-objective optimization model based on constraints. The preset distribution network control switching conditions include at least communication interruption. Based on the optimal global control strategy, multiple sub-regions are controlled to execute the first control command, and the sub-regions include user distributed photovoltaic power generation. If the operation of the distribution network meets the preset distribution network control switching conditions, the distribution network is controlled to execute a distributed control strategy, and the optimal local control strategy corresponding to each sub-region is obtained to achieve load balance across the entire network. Based on the optimal local control strategy, the corresponding sub-region is controlled to execute the second control command.

2. The method for coordinated control of power generation, load, and storage in a distribution network based on distributed power sources according to claim 1, characterized in that, The preset distribution network control switching conditions also include grid load fluctuation exceeding a load fluctuation threshold and grid load exceeding a load threshold; the communication interruption includes communication delay exceeding a delay threshold and network packet loss rate exceeding a packet loss rate threshold. The expression for the power grid load fluctuation exceeding the load fluctuation threshold is as follows: |P load (t)-P load (t-1)|>ΔP threshold ; In the formula, P load (t) and P load (t-1) represent the grid loads at times t and t-1, respectively, ΔP threshold The preset load fluctuation threshold; The expression for the communication delay exceeding the delay threshold is: Delay(t)>ΔDelay threshold ; In the formula, Delay(t) is the communication delay at the current time, and ΔDelay threshold The preset maximum communication delay threshold; The expression for the network packet loss rate exceeding the packet loss rate threshold is: In the formula, P lost P represents the number of packets lost. total For the total number of cases, ΔP rate This is the packet loss rate threshold; The expression for the power grid load exceeding the load threshold is: P load (t)>α·P max ; In the formula, P load (t) represents the current grid load, P max α represents the maximum load of the power grid, and α is the load ratio.

3. The method for coordinated control of power generation, load, and storage in a distribution network based on distributed power sources according to claim 1, characterized in that, The expression for the first multi-objective optimization model is: In the formula, C gen,i Let C be the resource cost of the i-th region. stor,i For the charging and discharging cost of energy storage systems, C solar,i For the costs associated with users generating electricity from solar panels, L bal The cost of load balancing, in order to balance the grid load with generated electricity, R gen λ1 and λ2 are the adjustment parameters, representing the utilization rate of renewable energy. The expression for the constraint condition is: In the formula, P load,i For the load demand of the i-th region, P gen,i For power generation, P stor,i For the power of energy storage devices, P solar,i This refers to the output power of distributed photovoltaic power generation. The expression for the safety constraints of the energy storage device is: In the formula, SOC i (t) represents the battery state of the energy storage device, SOC. min To minimize the energy stored by energy storage devices, SOC max This represents the maximum energy storage capacity of the energy storage device.

4. The method for coordinated control of power generation, load, and storage in a distribution network based on distributed power sources according to claim 1, characterized in that, The distributed regulation and control strategy includes a second multi-objective optimization model based on the constraints, and the expression of the second multi-objective optimization model is: minf i (x i )=C stor,i +C solar,i ; In the formula, C stor,i For the cost of energy storage equipment, C solar,i For the relevant costs of distributed photovoltaic power generation, x i Let be the scheduling decision variable for the i-th sub-region.

5. The method for coordinated control of power generation, load, and storage in a distribution network based on distributed power sources according to claim 1, characterized in that, If the distribution network operation meets the preset distribution network control switching conditions, then the distribution network is controlled to execute a distributed control strategy, and the optimal local control strategy corresponding to each sub-region is obtained, specifically including: Initialize the decision variables, dual variables, and global decision variables for each sub-region; Real-time acquisition of local data in sub-regions; The gradient is calculated based on the local data and the second multi-objective optimization model: In the formula, α is the learning rate. Let z be the gradient of the second multi-objective optimization model for the i-th sub-region. k For the global scheduling target, Let i be the dual variable of the i-th subregion. Let be the decision variable for the i-th sub-region in the k-th iteration, and ρ be the penalty factor; The sub-region is initially scheduled based on the gradient; The scheduling result is obtained based on the preliminary scheduling; The scheduling results are shared with other sub-regions, and the optimal local control strategy is obtained by solving the local data and the second multi-objective optimization model.

6. The method for coordinated control of power generation, load, and storage in a distribution network based on distributed power sources according to any one of claims 1 to 5, characterized in that, Also includes: Under the premise of implementing a centralized control strategy for the distribution network, a third multi-objective optimization model is constructed based on the distribution network operation data: In the formula, C gen,i Let C be the power generation cost of the i-th sub-region. stor,i The charging and discharging cost of the energy storage device in the i-th sub-region, C solar,i L represents the relevant costs of distributed photovoltaic power generation in the i-th region. bal R is a load balance constraint term. gen For the overall utilization rate of renewable energy, 1-R gen The term promotes the system to maximize the use of renewable energy, where λ1 and λ2 are weighting coefficients. Solve the third multi-objective optimization model to obtain the optimal distributed photovoltaic power generation dispatch amount; If the optimal distributed photovoltaic power generation call quantity is greater than zero, then the distributed photovoltaic power generation feedback is invoked.

7. The method for coordinated control of power generation, load, and storage in a distribution network based on distributed power sources according to any one of claims 1 to 5, characterized in that, Also includes: In implementing a distributed control strategy for the distribution network, a dynamic electricity price is constructed based on the network-wide load forecast and the distribution network operation data. p(t)=αL(t)+βR(t)+γ·SOC(t); p(t) is the electricity price at time t, L(t) is the total grid load forecast at time t, R(t) is the renewable energy generation at time t, SOC(t) is the charging status of electric vehicles or energy storage systems, α is the load weighting coefficient, β is the renewable energy weighting coefficient, and γ is the energy storage status weighting coefficient. A power grid scenario is constructed based on the dynamic electricity price signal. The power grid scenario includes peak load periods, low single load periods with sufficient sunshine, periods with high output of other renewable energy sources and low power generation of other renewable energy sources. Based on the aforementioned power grid scenario, electricity prices may be increased or decreased to incentivize distributed photovoltaic power generation users to feed electricity back to the grid.

8. The method for coordinated control of power generation, load, and storage in a distribution network based on distributed power sources according to claim 7, characterized in that, Also includes: Establish reward rules: R reward (t)=δ·P feedback (t)·η(t)·T(t); In the formula, R reward P(t) represents the user's reward at time t. feedback η(t) represents the electrical energy fed back to the grid by the user, η(t) represents the grid load factor, T(t) represents the duration of user participation in feedback, and δ represents the reward factor, which is the basic amount for adjusting the reward. The system calculates and publishes real-time electricity prices and reward rules based on the current power grid status and the dynamic electricity price, so that users can adjust their behavior of feeding electricity back to the power grid according to the received real-time electricity price and reward rules; Analyze the deviation between users' behavior in reporting electricity consumption to the power grid and the expected target; If a user's discharge volume is insufficient, the electricity price and rewards will be increased; If a user discharges a sufficient amount of electricity, the electricity price will be reduced and a reward will be given.

9. An electronic device, characterized in that, include: At least one processor; A memory that is communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the distributed power grid source-load-storage coordinated control method as described in any one of claims 1 to 8.

10. A medium, characterized in that, The system contains a computer program that, when executed by a processor, implements the power generation, load, and storage coordination control method for a distribution network based on distributed power sources as described in any one of claims 1 to 8.

Images