Energy scheduling method and system based on LSTM-PSO and virtual power plant

The LSTM-PSO method predicts power load and renewable energy generation in virtual power plants, and combines PSO optimization algorithm to update the state of energy storage units in real time, solving the rationality of power scheduling and improving the stability and economic benefits of power grid operation.

CN119029835APending Publication Date: 2024-11-26HUADIAN (ZHEJIANG) ENERGY SALES CO LTD +1
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410861562.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-28
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The prior art is difficult to effectively improve the rationality of electric energy scheduling, especially when facing the instability of renewable energy generation and unpredictability of load demand, resulting in unstable grid operation.

Method used

Using the energy scheduling method based on LSTM-PSO, the power load and renewable energy generation data of virtual power plants are collected, and the LSTM network is used to predict and PSO particle swarm optimization algorithm is used to find the lowest scheduling solution for virtual power plants, and the energy storage unit is updated in real time. State of charge.

Benefits of technology

It improves the accuracy and flexibility of electric energy scheduling, enhances the adaptability to renewable energy fluctuations, extends the life of energy storage units, and realizes energy scheduling with the lowest cost of virtual power plant scheduling.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119029835A_ABST
    Figure CN119029835A_ABST
Patent Text Reader

Abstract

The invention relates to an LSTM-PSO-based energy scheduling method and system and a virtual power plant, and the method comprises the steps: collecting the power load data and renewable energy power generation data of the virtual power plant; setting the energy scheduling target of the virtual power plant as the lowest scheduling cost of the virtual power plant; based on the power load data and the renewable energy power generation data, obtaining a power load predicted value and a renewable energy power generation predicted value through an LSTM network model; based on the power load predicted value and the renewable energy power generation predicted value, a target energy scheduling scheme with the lowest scheduling cost is obtained through a PSO particle swarm optimization algorithm; and updating the charge state of the energy storage unit by using the charge state prediction value. According to the invention, the scheduling scheme with the lowest virtual power plant scheduling cost as the target is realized, the economic benefit of the VPP is improved, the energy storage unit is updated in real time by using the state-of-charge predicted value, the service life of the energy storage unit can be effectively prolonged, the scheduling precision of the energy storage unit can be effectively improved, and the problem of how to improve the reasonability of electric energy scheduling is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of power dispatching, and in particular to an energy dispatching method, system and virtual power plant based on LSTM-PSO. Background Art

[0002] In recent years, with the rapid development of renewable energy technologies such as wind and solar energy, distributed generation (DG) has played an increasingly important role in the power grid. However, the power generation of these clean energy sources has brought unprecedented challenges to the safe operation of the power grid and the stability of power supply due to its inherent instability, such as randomness and intermittency affected by weather changes.

[0003] Electric energy dispatch is usually based on prediction and non-real-time optimization, that is, energy dispatch decisions rely on the prediction of load demand and distributed power output. However, in actual operation, it is extremely difficult to achieve high-accuracy predictions due to the unpredictability of loads and the instability of distributed power sources. In addition to improving prediction accuracy, developing real-time online algorithms to dynamically implement electric energy dispatch is also an effective solution. Real-time electric energy dispatch reduces the reliance on accurate predictions. But it also leads to the requirement for real-time dispatch.

[0004] Currently, no effective solution has been proposed for the problem of how to improve the rationality of power scheduling in related technologies. Summary of the invention

[0005] The embodiments of the present application provide an energy scheduling method, system and virtual power plant based on LSTM-PSO to at least solve the problem of how to improve the rationality of power scheduling in the related art.

[0006] In a first aspect, an embodiment of the present application provides an energy scheduling method based on LSTM-PSO, the method comprising:

[0007] Collect power load data and renewable energy generation data generated during the operation of the virtual power plant;

[0008] Setting the energy dispatch target of the virtual power plant to minimize the dispatch cost of the virtual power plant;

[0009] Based on the power load data and the renewable energy generation data, the power load forecast value and the renewable energy generation forecast value under the current scheduling period are predicted through the LSTM network model;

[0010] Based on the power load forecast value and the renewable energy power generation forecast value, a target energy scheduling solution with the lowest scheduling cost in the virtual power plant is found through a PSO particle swarm optimization algorithm;

[0011] Based on the target energy scheduling scheme, a state of charge prediction value of the energy storage unit in the current scheduling period is obtained, wherein the state of charge prediction value is used to update the state of charge of the energy storage unit.

[0012] In some embodiments, setting the energy dispatch target of the virtual power plant to minimize the virtual power plant dispatch cost includes:

[0013] Through the objective function minCost(P d , P ess , P grid , M)=C d +C ess +C grid , the energy dispatch target of the virtual power plant is set to the lowest virtual power plant dispatch cost, where P d is the output of the thermal power generation unit, P ess is the charge / discharge power of the energy storage unit, P grid is the exchange power between the virtual power plant and the external power grid, M is the number of scheduling periods for energy scheduling, C d is the power generation cost of the thermal power generation unit, C ess is the dispatch cost of the energy storage unit, C grid The cost of purchasing electricity from the external grid.

[0014] In some embodiments, the method comprises:

[0015] The power generation cost of the thermal power generation unit, the dispatching cost of the energy storage unit and the cost of purchasing electricity from the external power grid are calculated through the dispatching cost calculation formula. The dispatching cost calculation formula is:

[0016]

[0017] Among them, P d (k) is the output of the thermal power generation unit in the scheduling period k, P ess (k) is the charge / discharge power of the energy storage unit in the scheduling period k, P grid (k) is the exchange power between the virtual power plant and the external power grid in the dispatch period k, c d1 and c d2 are the primary cost coefficient and secondary cost coefficient of the thermal power generation unit, C ess is the cost coefficient of the energy storage unit, and EP(k) is the time-of-use electricity price.

[0018] In some embodiments, based on the power load data and the renewable energy generation data, the power load forecast value and the renewable energy generation forecast value under the current scheduling period are predicted by the LSTM network model, including:

[0019] Based on the power load data, the power load forecast value under the current scheduling period is predicted through the LSTM network model;

[0020] Based on the renewable energy power generation data, the renewable energy power generation forecast value in the current scheduling period is predicted through the LSTM network model, wherein the renewable energy power generation forecast value includes the photovoltaic power generation forecast value and the wind power generation forecast value.

[0021] In some embodiments, based on the power load forecast value and the renewable energy power generation forecast value, finding a target energy scheduling solution with the lowest scheduling cost in the virtual power plant through a PSO particle swarm optimization algorithm includes:

[0022] Based on the power load forecast value and the renewable energy power generation forecast value, the PSO particle swarm optimization algorithm is used to find the target energy scheduling solution with the lowest scheduling cost in the virtual power plant:

[0023] P d,best =[P d (t), P d (t+1), ..., P d (N)]

[0024] P ess,best =[P ess (t), P ess (t+1), ..., P ess (N)]

[0025] P grid,best =[P grid (t), P grid (t+1), .., P grid (N)]

[0026] Among them, P d,best is the target decision value of the thermal power generation unit output, P ess,best is the target decision value of the charging / discharging power of the energy storage unit, P grid,best is the target decision of the exchange power between the virtual power plant and the external grid, P d (t), P d (t+1), ..., P d (N) is the output of the thermal power generation unit in each scheduling period, P ess (t), P ess (t+1), ..., P ess (N) is the charge / discharge power of the energy storage unit in each scheduling period, P grid (t), P grid (t+1), ..., P grid(N) is the exchange power between the virtual power plant and the external power grid in each scheduling period.

[0027] In some embodiments, based on the target energy scheduling scheme, the charge state prediction value of the energy storage unit in the current scheduling period is obtained, which includes:

[0028] Based on the target energy scheduling scheme, the charging / discharging power of the energy storage unit in the current scheduling period is calculated by a power balance formula;

[0029] Based on the charge / discharge power of the energy storage unit in the current scheduling period, the energy storage SOC update formula The predicted value of the state of charge of the energy storage unit in the current scheduling period is obtained, where P ess (t) is the charge / discharge power of the energy storage unit in the scheduling period t, η charge is the charge / discharge efficiency of the energy storage unit, ES max is the maximum capacity of the energy storage unit, S oc (t) is the charge state of the energy storage unit in the scheduling period t, S oc (t-1) is the charge state of the energy storage unit in the previous scheduling period t-1.

[0030] In some embodiments, based on the target energy scheduling scheme, the charge / discharge power of the energy storage unit in the current scheduling period is calculated by the power balance formula, including:

[0031] Based on the target energy scheduling scheme, the power balance formula P wind (t)+P pv (t)+e d (t)+P ess (t)+P grid (t) = P load (t), calculate the charging / discharging power of the energy storage unit in the current scheduling period, where P wind (t) is the output value of the wind power generation unit in the dispatching period t, P pv (t) is the output value of the photovoltaic power generation unit in the scheduling period t, P d (t) is the output of the thermal power generation unit in the dispatching period t, P ess (t) is the charging / discharging power of the energy storage unit in the scheduling period t, P grid (t) is the exchange power between the virtual power plant and the external power grid in the dispatch period t, P load (t) is the power load during the dispatch period t.

[0032] In some embodiments, collecting power load data and renewable energy generation data generated during the operation of the virtual power plant includes:

[0033] The power load data and renewable energy generation data generated by the virtual power plant in each scheduling period are collected in sequence, that is, the power load data and the renewable energy generation data are time series data.

[0034] In a second aspect, an embodiment of the present application provides an energy scheduling system based on LSTM-PSO, the system being used to execute the method described in any one of the first aspects above, the system comprising a data acquisition module, a target setting module and a scheduling optimization module;

[0035] The data acquisition module is used to collect power load data and renewable energy power generation data generated during the operation of the virtual power plant;

[0036] The target setting module is used to set the energy scheduling target of the virtual power plant to minimize the scheduling cost of the virtual power plant;

[0037] The scheduling optimization module is used to predict the power load forecast value and the renewable energy generation forecast value in the current scheduling period according to the power load data and the renewable energy generation data through the LSTM network model; based on the power load forecast value and the renewable energy generation forecast value, through the PSO particle swarm optimization algorithm, find the target energy scheduling scheme with the lowest scheduling cost of the virtual power plant; based on the target energy scheduling scheme, obtain the charge state prediction value of the energy storage unit in the current scheduling period, wherein the charge state prediction value is used to update the charge state of the energy storage unit.

[0038] In the third aspect, an embodiment of the present application provides an energy scheduling virtual power plant based on LSTM-PSO, wherein the virtual power plant is used to execute the method described in any one of the first aspects above, and the virtual power plant includes a thermal power generation unit, a renewable energy power generation unit, an energy storage unit, a power load unit and an energy scheduling unit, wherein the renewable energy power generation unit includes a wind power generation unit and a photovoltaic power generation unit.

[0039] Compared with the related art, the embodiment of the present application provides an energy scheduling method, system and virtual power plant based on LSTM-PSO, wherein the method collects the power load data and renewable energy power generation data generated by the virtual power plant during operation; sets the energy scheduling target of the virtual power plant to the lowest scheduling cost of the virtual power plant; based on the power load data and renewable energy power generation data, the LSTM network model is used to predict the power load forecast value and renewable energy power generation forecast value in the current scheduling period; based on the power load forecast value and renewable energy power generation forecast value, the PSO particle swarm optimization algorithm is used to find the target energy scheduling solution with the lowest scheduling cost of the virtual power plant; based on The target energy scheduling plan obtains the charge state prediction value of the energy storage unit in the current scheduling period, wherein the charge state prediction value is used to update the charge state of the energy storage unit, realizing the energy scheduling plan with the lowest virtual power plant scheduling cost as the goal. The uncontrollable power load and renewable energy generation are predicted in time series by LSTM, which greatly improves the prediction accuracy. At the same time, the plan for each scheduling period is obtained in real time by PSO, which improves the flexibility of virtual power plant energy scheduling and adaptability to renewable energy fluctuations. In addition, the use of the charge state prediction value to update the charge state of the energy storage unit in real time can effectively improve the life of the energy storage unit and solve the problem of how to improve the rationality of power scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0041] Figure 1 is a flowchart of the steps of the energy scheduling method based on LSTM-PSO according to an embodiment of the present application;

[0042] Figure 2 is a schematic diagram of a long short-term memory network model provided according to an embodiment of the present application;

[0043] Figure 3 is a structural block diagram of an energy scheduling system based on LSTM-PSO according to an embodiment of the present application;

[0044] Figure 4 It is a schematic diagram of the internal structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. Based on the embodiments provided in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0046] Obviously, the drawings described below are only some examples or embodiments of the present application. For ordinary technicians in this field, the present application can also be applied to other similar scenarios based on these drawings without creative work. In addition, it can also be understood that although the efforts made in this development process may be complicated and lengthy, for ordinary technicians in this field related to the content disclosed in this application, some changes in design, manufacturing or production based on the technical content disclosed in this application are just conventional technical means, and should not be understood as insufficient content disclosed in this application.

[0047] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0048] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the" and the like involved in this application do not indicate a quantitative limitation, and may represent the singular or plural. The terms "include", "comprise", "have" and any of their variations involved in this application are intended to cover non-exclusive inclusions; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units that are not listed, or may also include other steps or units inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0049] The embodiment of the present application provides an energy scheduling method based on LSTM-PSO. Figure 1 is a flowchart of the steps of the energy scheduling method based on LSTM-PSO according to an embodiment of the present application, such as Figure 1 As shown, the method comprises the following steps:

[0050] Step S102, collecting power load data and renewable energy generation data generated during the operation of the virtual power plant;

[0051] Specifically, step S102 collects the power load data and renewable energy generation data generated by the virtual power plant in each scheduling period in sequence, that is, the power load data and renewable energy generation data are time series data.

[0052] In step S102, preferably, the current scheduling period t=1 is set, and 24 hours a day is divided into N scheduling periods, each scheduling period is 15 minutes, then N=96. Before optimizing the scheduling, the state of charge (SOC) of the energy storage unit (ESS) is usually initialized to 0.5, which is updated in each scheduling period and affects the optimization decision.

[0053] It should be noted that the core advantage of Virtual Power Plant (VPP) lies in its ability to efficiently integrate the dispersed power generation units, energy storage units (ESS) and controllable loads within its jurisdiction to form a coordinated grid access entity, thereby improving economic benefits and system operation efficiency. In the operation of VPP, due to the integration of a large number of distributed power sources, its power generation inevitably presents intermittent, volatile and unpredictable characteristics. In order to effectively manage these uncertainties, in addition to the traditional methods of improving prediction accuracy, this paper proposes an innovative VPP energy scheduling method based on Long Short-Term Memory (LSTM) and Particle Swarm Optimization (PSO) on the basis of existing research. This method focuses on maximizing the optimization control effect within a limited time range in the future, and introduces a feedback correction mechanism to effectively reduce the impact of prediction errors on energy scheduling and improve the accuracy of energy scheduling. The energy scheduling scheme with the lowest virtual power plant scheduling cost as the goal not only improves the economic benefits of VPP, but also reduces the impact on the main power grid. In other words, VPPs can dispatch energy with greater flexibility and precision, optimize grid operation, enhance economic benefits, and reduce potential threats to external grid stability.

[0054] Step S104, setting the energy dispatch target of the virtual power plant to minimize the dispatch cost of the virtual power plant;

[0055] Step S104 specifically, through the objective function minCost(P d , P ess , P grid , M)=C d +C ess +C grid , the energy dispatch target of the virtual power plant is set to the minimum dispatch cost of the virtual power plant, where P d is the output of the thermal power generation unit, P ess is the charge / discharge power of the energy storage unit, P grid is the exchange power between the virtual power plant and the external power grid, M is the number of scheduling periods for energy scheduling, C d is the power generation cost of the thermal power generation unit, C ess is the dispatch cost of the energy storage unit, C grid The cost of purchasing electricity from the external grid.

[0056] Preferably, step S104 uses the scheduling cost calculation formula:

[0057]

[0058] Calculate the power generation cost C of the thermal power generation unit d , the dispatch cost of the energy storage unit C ess and the cost of purchasing electricity from the external grid C grid , where P d (k) is the output of the thermal power generation unit in the scheduling period k, P ess (k) is the charge / discharge power of the energy storage unit in the scheduling period k, P grid (k) is the exchange power between the virtual power plant and the external power grid in the dispatch period k, c d1 and c d2 are the primary cost coefficient and secondary cost coefficient of the thermal power generation unit, C ess is the cost coefficient of the energy storage unit, and EP(k) is the time-of-use electricity price.

[0059] It should be noted that the energy dispatch scheme with the goal of minimizing the virtual power plant dispatch cost not only improves the economic benefits of the VPP, but also reduces the impact on the main power grid.

[0060] Step S106, based on the power load data and the renewable energy generation data, the power load forecast value and the renewable energy generation forecast value in the current scheduling period are predicted through the LSTM network model;

[0061] Specifically, in step S106, based on the power load data, the power load forecast value in the current scheduling period is predicted through the LSTM network model;

[0062] Based on the renewable energy power generation data, the LSTM network model is used to predict the renewable energy power generation forecast value under the current scheduling period, where the renewable energy power generation forecast value includes the photovoltaic power generation forecast value and the wind power generation forecast value.

[0063] It should be noted that Figure 2 is a schematic diagram of a long short-term memory network model provided according to an embodiment of the present application, such as Figure 2 As shown in the figure, the LSTM network is a type of recurrent neural network (RNN). It mainly solves the gradient explosion and gradient vanishing problems encountered when training RNN. These problems will cause the feature information of the earlier time step to be unable to be transferred to the later time step, causing the network to lose important information. The core feature of LSTM is that it establishes an information flow between all time steps and controls the retention and forgetting of information at each time step through a gating mechanism. In the LSTM recurrent unit, a complex calculation formula is used, among which the most critical core structure is called Cell State. e ). The cell state is the information flow throughout the LSTM unit. Figure 2 C t-1 to C t The straight arrow flow is represented by .

[0064] The LSTM unit mainly consists of the following structure:

[0065] Forget Gate.

[0066] In the LSTM unit, we first decide what information the cell should forget. We take the hidden state ht- of the previous unit and the input sequence x of the current unit. t Input the forget gate and generate the output vector f after passing the sigmoid function t , whose components are in the range [0, 1]. When the value of the component in the vector is 0, it means that C t-1 The information of is not retained at all; when it is 1, it means C t-1 The information is fully preserved:

[0067] f t =σ(W f ·[h t-1 , x t ]+b f )

[0068] Among them, σ() is the sigmoid activation function in the neural network, W f Represents the weight parameter.

[0069] Input Gate.

[0070] Similar to the forget gate, the input gate is also based on h t-1 and x t Calculate the information of the input cell state. However, the calculation of the input gate is more complicated and includes two parts: the first part uses the tanh function to convert h t-1 and x t Transformed into a new information vector The second part is the same as the forget gate, generating the vector i through the sigmoid function t , whose components take values ​​in [0, 1]. and i t Information needed to update the cell state:

[0071]

[0072] i t =σ(W i ·[h t-1 , x t ]+b i )

[0073] Among them, tanh represents the activation function in the neural network, W c and W i represents the weight parameter, b c and b i Represents the bias parameter.

[0074] Update Gate.

[0075] After passing through the forget gate and input gate, the discarded and updated information is determined. Then the cell state is updated to obtain C t :

[0076]

[0077] Output Gate.

[0078] After updating the cell state, the input h is first processed by the sigmoid function t-1 and x t , get the judgment condition vector C t Then make the cell state C t After the tanh function, we get a vector value with each component value of [-1, 1]. This vector is the same as o t Multiply them together to get the output h of the LSTM unit t :

[0079] o t =σ(W o ·[h t-1 , x t ]+b o )

[0080] h t =o t *tanh(C t )

[0081] Among them, W o represents the weight coefficient, b o Represents the bias coefficient. LSTM implements a time series data processing model with long short-term memory through its internal gating mechanism and model loop structure.

[0082] The deep learning capability of the LSTM network can provide VPP with more accurate load and power generation forecasts, which lays a solid foundation for achieving better energy dispatch and improving system economic benefits. In addition, the introduction of LSTM also enhances VPP's ability to manage the volatility and uncertainty of renewable energy, thereby optimizing the operating efficiency and reliability of the entire power system.

[0083] Step S108, based on the power load forecast value and the renewable energy power generation forecast value, a PSO particle swarm optimization algorithm is used to find a target energy scheduling solution with the lowest scheduling cost in the virtual power plant;

[0084] Specifically, step S108 is to find the target energy scheduling scheme with the lowest scheduling cost in the virtual power plant through the PSO particle swarm optimization algorithm based on the power load forecast value and the renewable energy power generation forecast value:

[0085] P d,best =[P d (t), P d (t+1), ..., P d (N)]

[0086] P ess,best =[P ess (t), P ess (t+1), ..., P ess (N)]

[0087] P grid,best =[P grid (t), P grid (t+1), .., P grid (N)]

[0088] Among them, P d,best is the target decision value of the thermal power generation unit output, P ess,best is the target decision value of the charging / discharging power of the energy storage unit, P grid , best is the target decision of the exchange power between the virtual power plant and the external grid, P d (t), Pd(t+1), ..., P d (N) is the output of the thermal power generation unit in each scheduling period, P ess (t), P ess (t+1), ..., P ess (N) is the charge / discharge power of the energy storage unit in each scheduling period, P grid (t), P grid (t+1), ..., P grid (N) is the exchange power between the virtual power plant and the external power grid in each scheduling period.

[0089] It should be noted that step S108 specifically calculates the P in each scheduling period step by step. d , P ess and P grid The optimal scheduling value can realize efficient and coordinated optimization scheduling of energy resources within VPP.

[0090] It should be further explained that the PSO algorithm is used to find the optimal energy scheduling solution based on the LSTM prediction. The PSO algorithm (Particle Swarm Optimization) assumes that there are n massless and volumeless particles in N-dimensional space, and the vector x i =(x i1 , x i2 …, x iN ) represents their position, vector v i =(v i1 , v i2 …, v iN ) represents their moving speed. Each particle has a fitness value determined by its objective function, and the particle seeks its optimal position p during movement. i =(p i1 , p i2 …, p iN ) and the optimal position g of the neighboring particles i =(g i1 , g i2 …, g iN ). Particles get new position parameters by continuously moving and updating iteratively. The formulas for updating their speed and position are as follows:

[0091] v ij (t+1)=ωv ij (t)+c1r1(p iN (t)-x iN (t))+c2r2(g iN (t)-x iN (t))

[0092] x ij (t+1)=ωv ij (t)+v ij (t+1)

[0093] Where: ω is the inertia parameter. When ω>1, the particle motion accelerates. When ω<1, the particle motion decelerates. The value of ω decreases from 0.9 to 0.4. c1 and c2 are learning parameters. Generally, c1=c2=2. r1 and r2 are random numbers in [0, 1]. ij (t+1) is the position parameter of particle i in the jth dimension iteration t+1 times, v ij (t+1) is the velocity parameter of particle i at the t+1th iteration of the i-th dimension.

[0094] Step S110, based on the target energy scheduling scheme, obtain a predicted value of the state of charge of the energy storage unit in the current scheduling period, wherein the predicted value of the state of charge is used to update the state of charge of the energy storage unit.

[0095] Step S110 specifically includes the following steps:

[0096] Step S1101, based on the target energy scheduling scheme, the charging / discharging power of the energy storage unit in the current scheduling period is calculated by a power balance formula;

[0097] Step S1101 Specifically, based on the target energy scheduling scheme, the power balance formula P wind (t)+P pv (t)+P d (t)+P ess (t)+P grid (t) = P load (t), calculate the charging / discharging power of the energy storage unit in the current scheduling period, where P wind (t) is the output value of the wind power generation unit in the dispatching period t, P pv (t) is the output value of the photovoltaic power generation unit in the scheduling period t, P d (t) is the output of the thermal power generation unit in the dispatching period t, P ess (t) is the charging / discharging power of the energy storage unit in the scheduling period t, P grid (t) is the exchange power between the virtual power plant and the external power grid in the dispatch period t, P load (t) is the power load during the dispatch period t.

[0098] Step S1102: Based on the charge / discharge power of the energy storage unit in the current scheduling period, the energy storage SOC is updated by the formula The predicted value of the state of charge of the energy storage unit in the current scheduling period is obtained, where P ess (t) is the charge / discharge power of the energy storage unit in the scheduling period t, η charge is the charge / discharge efficiency of the energy storage unit, ES max is the maximum capacity of the energy storage unit, S oc (t) is the charge state of the energy storage unit in the scheduling period t, S oc (t-1) is the charge state of the energy storage unit in the previous scheduling period t-1.

[0099] It should be noted that in step S110, after the state of charge of the energy storage unit is updated using the state of charge prediction value, during the iteration process of the above-mentioned particle swarm optimization (PSO) algorithm, the SOC value of the energy storage unit is continuously monitored to see whether it exceeds its preset threshold, and the SOC value is adjusted using the actual measured load, wind power and photovoltaic power generation, and the actual charging and discharging power of the energy storage system is fine-tuned to form a closed-loop optimization to improve the optimization accuracy.

[0100] Through the above steps in the embodiment of the present application, an energy scheduling plan with the goal of minimizing the virtual power plant scheduling cost is achieved. The uncontrollable power load and renewable energy generation are predicted in time series by LSTM, which greatly improves the prediction accuracy. At the same time, the plan for each scheduling period is derived in real time by PSO, which improves the flexibility of virtual power plant energy scheduling and adaptability to renewable energy fluctuations. In addition, the state of charge of the energy storage unit is updated in real time using the state of charge prediction value, which can effectively improve the life of the energy storage unit and solve the problem of how to improve the rationality of power scheduling.

[0101] It should be noted that the steps shown in the above process or the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0102] The embodiment of the present application provides an energy scheduling system based on LSTM-PSO. Figure 3 is a structural block diagram of an energy scheduling system based on LSTM-PSO according to an embodiment of the present application, such as Figure 3 As shown, the system includes a data acquisition module, a target setting module and a scheduling optimization module;

[0103] A data acquisition module is used to collect power load data and renewable energy generation data generated during the operation of the virtual power plant;

[0104] A target setting module, used to set the energy dispatch target of the virtual power plant to minimize the dispatch cost of the virtual power plant;

[0105] The scheduling optimization module is used to predict the power load forecast value and renewable energy generation forecast value in the current scheduling period based on the power load data and renewable energy generation data through the LSTM network model; based on the power load forecast value and renewable energy generation forecast value, the PSO particle swarm optimization algorithm is used to find the target energy scheduling plan with the lowest scheduling cost in the virtual power plant; based on the target energy scheduling plan, the charge state prediction value of the energy storage unit in the current scheduling period is obtained, wherein the charge state prediction value is used to update the charge state of the energy storage unit.

[0106] Through the data acquisition module, target setting module and scheduling optimization module in the embodiment of the present application, an energy scheduling plan with the goal of minimizing the scheduling cost of the virtual power plant is achieved. The uncontrollable power load and renewable energy generation are predicted in time series by LSTM, which greatly improves the prediction accuracy. At the same time, the plan for each scheduling period is derived in real time by PSO, which improves the flexibility of the virtual power plant's energy scheduling and its adaptability to renewable energy fluctuations. In addition, the state of charge of the energy storage unit is updated in real time using the state of charge prediction value, which can effectively increase the life of the energy storage unit and solve the problem of how to improve the rationality of power scheduling.

[0107] It should be noted that the above modules can be functional modules or program modules, and can be implemented by software or hardware. For modules implemented by hardware, the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0108] An embodiment of the present application provides an energy scheduling virtual power plant based on LSTM-PSO, which includes a thermal power generation unit, a renewable energy power generation unit, an energy storage unit, a power load unit and an energy scheduling unit, wherein the renewable energy power generation unit includes a wind power generation unit and a photovoltaic power generation unit.

[0109] This embodiment further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0110] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0111] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be described in detail here.

[0112] In addition, in combination with the energy scheduling method based on LSTM-PSO in the above embodiment, the embodiment of the present application can provide a storage medium for implementation. The storage medium stores a computer program; when the computer program is executed by the processor, any one of the energy scheduling methods based on LSTM-PSO in the above embodiment is implemented.

[0113] In one embodiment, a computer device is provided, which may be a terminal. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, an energy scheduling method based on LSTM-PSO is implemented. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0114] In one embodiment, Figure 4 is a schematic diagram of the internal structure of an electronic device according to an embodiment of the present application, such as Figure 4 As shown, an electronic device is provided, which may be a server, and its internal structure diagram may be as shown in Figure 4 As shown. The electronic device includes a processor, a network interface, an internal memory and a non-volatile memory connected through an internal bus, wherein the non-volatile memory stores an operating system, a computer program and a database. The processor is used to provide computing and control capabilities, the network interface is used to communicate with an external terminal through a network connection, the internal memory is used to provide an environment for the operation of the operating system and the computer program, the computer program is executed by the processor to implement an energy scheduling method based on LSTM-PSO, and the database is used to store data.

[0115] Those skilled in the art will understand that Figure 4 The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the electronic device to which the scheme of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0116] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), 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).

[0117] Those skilled in the art should understand that the technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0118] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. An energy scheduling method based on LSTM-PSO, characterized in that: The method comprises: Collect power load data and renewable energy generation data generated during the operation of the virtual power plant; Setting the energy dispatch target of the virtual power plant to minimize the dispatch cost of the virtual power plant; Based on the power load data and the renewable energy generation data, the power load forecast value and the renewable energy generation forecast value under the current scheduling period are predicted through the LSTM network model; Based on the power load forecast value and the renewable energy power generation forecast value, a target energy scheduling solution with the lowest scheduling cost in the virtual power plant is found through a PSO particle swarm optimization algorithm; Based on the target energy scheduling scheme, a state of charge prediction value of the energy storage unit in the current scheduling period is obtained, wherein the state of charge prediction value is used to update the state of charge of the energy storage unit.

2. The method according to claim 1, characterized in that Setting the energy dispatch target of the virtual power plant to the lowest virtual power plant dispatch cost includes: Through the objective function minCost(P d , P ess , P grid , M)=C a +C ess +C grid , the energy dispatch target of the virtual power plant is set to the lowest virtual power plant dispatch cost, where P d is the output of the thermal power generation unit, P ess is the charge / discharge power of the energy storage unit, P grid is the exchange power between the virtual power plant and the external power grid, M is the number of scheduling periods for energy scheduling, C d is the power generation cost of the thermal power generation unit, C ess is the dispatch cost of the energy storage unit, C grid The cost of purchasing electricity from the external grid.

3. The method according to claim 2, characterized in that The method comprises: The power generation cost of the thermal power generation unit, the dispatching cost of the energy storage unit and the cost of purchasing electricity from the external power grid are calculated through the dispatching cost calculation formula. The dispatching cost calculation formula is: Among them, P d (k) is the output of the thermal power generation unit in the scheduling period k, P ess (k) is the charge / discharge power of the energy storage unit in the scheduling period k, P grid (k) is the exchange power between the virtual power plant and the external power grid in the dispatch period k, c d1 and c d2 are the primary cost coefficient and secondary cost coefficient of the thermal power generation unit, C ess is the cost coefficient of the energy storage unit, and EP(k) is the time-of-use electricity price.

4. The method according to claim 1, characterized in that: Based on the power load data and the renewable energy generation data, the power load forecast value and renewable energy generation forecast value under the current scheduling period are predicted through the LSTM network model, including: Based on the power load data, the power load forecast value under the current scheduling period is predicted through the LSTM network model; Based on the renewable energy power generation data, the renewable energy power generation forecast value in the current scheduling period is predicted through the LSTM network model, wherein the renewable energy power generation forecast value includes the photovoltaic power generation forecast value and the wind power generation forecast value.

5. The method according to claim 1, characterized in that Based on the power load forecast value and the renewable energy power generation forecast value, the target energy scheduling scheme with the lowest scheduling cost in the virtual power plant is found through the PSO particle swarm optimization algorithm, including: Based on the power load forecast value and the renewable energy power generation forecast value, the PSO particle swarm optimization algorithm is used to find the target energy scheduling solution with the lowest scheduling cost in the virtual power plant: P d,best =[P d (t),P d (t+1),...,P d (N)] P ess,best =[P ess (t),P ess (t+1),...,P ess (N)] P grid,best =[P grid (t),P grid (t+1),...,P grid (N)] Among them, P d,best is the target decision value of the thermal power generation unit output, P ess,best is the target decision value of the charging / discharging power of the energy storage unit, P grid,best is the target decision of the exchange power between the virtual power plant and the external grid, P d (t), P d (t+1), ..., P d (N) is the output of the thermal power generation unit in each scheduling period, P ess (t), P ess (t+1), ..., P ess (N) is the charge / discharge power of the energy storage unit in each scheduling period, P grid (t), P grid (t+1), ..., P grid (N) is the exchange power between the virtual power plant and the external power grid in each scheduling period.

6. The method according to claim 1, characterized in that Based on the target energy scheduling scheme, the predicted state of charge values ​​of the energy storage unit in the current scheduling period include: Based on the target energy scheduling scheme, the charging / discharging power of the energy storage unit in the current scheduling period is calculated by a power balance formula; Based on the charge / discharge power of the energy storage unit in the current scheduling period, the energy storage SOC update formula The predicted value of the state of charge of the energy storage unit in the current scheduling period is obtained, where P ess (t) is the charge / discharge power of the energy storage unit in the scheduling period t, η charge is the charge / discharge efficiency of the energy storage unit, ES max is the maximum capacity of the energy storage unit, S oc (t) is the charge state of the energy storage unit in the scheduling period t, S oc (t-1) is the charge state of the energy storage unit in the previous scheduling period t-1.

7. The method according to claim 6, characterized in that Based on the target energy scheduling scheme, the charging / discharging power of the energy storage unit in the current scheduling period is calculated by the power balance formula, including: Based on the target energy scheduling scheme, the power balance formula P wind (t)+P pv (t)+P d (t)+P ess (t)+P grid (t) = P load (t), calculate the charging / discharging power of the energy storage unit in the current scheduling period, where P wind (t) is the output value of the wind power generation unit in the dispatching period t, P pv (t) is the output value of the photovoltaic power generation unit in the scheduling period t, P d (t) is the output of the thermal power generation unit in the dispatching period t, P ess (t) is the charging / discharging power of the energy storage unit in the scheduling period t, P grid (t) is the exchange power between the virtual power plant and the external power grid in the dispatch period t, P load (t) is the power load during the dispatch period t.

8. The method according to claim 1, characterized in that The power load data and renewable energy generation data generated by the virtual power plant during operation include: The power load data and renewable energy generation data generated by the virtual power plant in each scheduling period are collected in sequence, that is, the power load data and the renewable energy generation data are time series data.

9. An energy scheduling system based on LSTM-PSO, characterized in that: The system is used to execute the method described in any one of claims 1 to 8, and the system includes a data acquisition module, a target setting module and a scheduling optimization module; The data acquisition module is used to collect power load data and renewable energy power generation data generated during the operation of the virtual power plant; The target setting module is used to set the energy scheduling target of the virtual power plant to minimize the scheduling cost of the virtual power plant; The scheduling optimization module is used to predict the power load forecast value and the renewable energy power generation forecast value in the current scheduling period according to the power load data and the renewable energy power generation data through the LSTM network model; based on the power load forecast value and the renewable energy power generation forecast value, find the target energy scheduling solution with the lowest scheduling cost in the virtual power plant through the PSO particle swarm optimization algorithm; Based on the target energy scheduling scheme, a state of charge prediction value of the energy storage unit in the current scheduling period is obtained, wherein the state of charge prediction value is used to update the state of charge of the energy storage unit.

10. An energy scheduling virtual power plant based on LSTM-PSO, characterized in that: The virtual power plant is used to execute the method described in any one of claims 1 to 8, and the virtual power plant includes a thermal power generation unit, a renewable energy power generation unit, an energy storage unit, a power load unit and an energy scheduling unit, wherein the renewable energy power generation unit includes a wind power generation unit and a photovoltaic power generation unit.

Citation Information

Cited By

  • Multi-type capacity application method and system of virtual power plant

    CN119358974A

  • A multi-type capacity application method and system for a virtual power plant

    CN119358974B