A power system dispatching framework method containing electric vehicle resources and a system thereof

Abstract

A kind of power system dispatching framework method and its system containing electric vehicle resource, according to the electric vehicle charging station in each region collected electric vehicle charging and discharging history information, the charging and discharging capacity of each electric vehicle charging station in region is forecasted in day ahead with charging station or electric vehicle load aggregator as unit;According to the basic load information of regional power grid, determine the regional power grid control target and control demand;According to the determined control demand, form electric vehicle day-ahead charging and discharging plan;Combined with the basic load information of regional power grid and the generated electric vehicle day-ahead charging and discharging plan in region, form power system day-ahead dispatching plan, check security;According to electric vehicle day-ahead charging and discharging plan and power system day-ahead dispatching plan, carry out rolling correction of electric vehicle charging and discharging plan, real-time scheduling.The electric vehicle is regarded as source side resource and integrated into original dispatching process in the level of dispatching in the application, and the distributed energy storage resource characteristics of large-scale electric vehicle can be better played.

Description

Technical Field

[0001] This invention belongs to the field of power system dispatching and operation, and specifically relates to a power system dispatching framework method and system that includes electric vehicle resources. Background Technology

[0002] Developing new energy vehicles is a strategic measure for the country to address climate change and promote energy transformation, and it is also an important part of end-use electricity substitution. It is estimated that by 2030, my country's electric vehicle fleet will reach 83 million vehicles, with an equivalent energy storage capacity of 5 billion kilowatt-hours. Electric vehicle charging demand will account for 6% to 7% of total electricity consumption, and peak charging load will account for 11% to 12% of the power grid load. The large-scale charging load of electric vehicles will have a significant impact on urban power grids, inevitably exacerbating prominent contradictions such as peak-valley differences, voltage deviations, and localized congestion in the power system. At the same time, the distributed energy storage characteristics of electric vehicles will provide abundant dispatchable resources for power grid peak shaving, voltage regulation, and renewable energy consumption.

[0003] With the increasing randomness and volatility of large-scale electric vehicles, the traditional scheduling and control mode that only considers large-scale electric vehicles as loads is no longer suitable for their access requirements, and there is a lack of effective multi-level collaborative scheduling methods. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide a power system dispatching framework method and system incorporating electric vehicle resources. This invention proposes a power system dispatching framework that integrates electric vehicles as source-side resources into the existing dispatching process at the dispatching level. This approach can better leverage the distributed energy storage characteristics of large-scale electric vehicles, collaboratively regulate grid load, provide ancillary services to the grid, and enhance the regional grid regulation capacity.

[0005] The present invention adopts the following technical solution:

[0006] This invention proposes a power system dispatching framework method incorporating electric vehicle resources, comprising the following:

[0007] Step 1: Obtain time-series historical information based on the charging and discharging historical information of electric vehicles collected from each electric vehicle charging station in the region, and make day-ahead forecasts of the charging and discharging volume of each electric vehicle charging station in the region, using electric vehicle charging stations or electric vehicle load aggregators as units.

[0008] Step 2: Determine the regional power grid control objectives and control requirements based on the regional power grid's basic load information;

[0009] Step 3: Based on the control requirements determined in Step 2, control the charging and discharging volume of each electric vehicle charging station in the region obtained from the day-ahead forecast in Step 1 to form a day-ahead charging and discharging plan for electric vehicles in the region.

[0010] Step 4: Combining the regional power grid basic load information from Step 2 and the regional electric vehicle daytime charging and discharging plan generated in Step 3, a daytime dispatch plan for the power system is formed, and its safety is verified.

[0011] Step 5: Based on the daytime charging and discharging plan of electric vehicles in the region and the daytime dispatching plan of the power system, collect and combine the charging and discharging information of electric vehicles in advance and in real time, and perform rolling correction of the daytime charging and discharging plan of electric vehicles to achieve real-time dispatching.

[0012] In step 1, the collected historical charging and discharging information of electric vehicles includes location data, average travel mileage, start charging time, estimated departure time, start SOC, and demand SOC.

[0013] In step 1, the time series historical information of electric vehicle charging and discharging history is stabilized using a difference algorithm.

[0014] The forecasting model used in the current forecast is ARIMA(p,d,q), where p, d, and q are the autoregressive order, differencing order, and moving average order, respectively; the historical time series information is denoted as Ori={T i}(i=1,2,…,N),T i This represents the time-series historical information of the i-th electric vehicle; N is the number of electric vehicles per unit area.

[0015] ARIMA(p,d,q) specifically means:

[0016] φ(B)(1-B) d t i =θ(B)ε t

[0017]

[0018] θ(B)=1-θ1B-θ2B2-…-θ q B q

[0019] In the formula, B is the shift operator, ε t To obey N(0,σ) 2 Error term of normal distribution; θ i Represents the autocorrelation coefficient of the corresponding time series, i = 1, 2, ..., q; φ(B) represents the moving average coefficient of the corresponding time series, j = 1, 2, ..., p; φ(B) represents the autoregressive model of the time series, and θ(B) represents the moving average model of the time series.

[0020] The enhanced Dickey-Fowler test is used to determine the range of values ​​for the autoregressive order, the difference order, and the moving average order in the prediction model. Then, the minimum information criterion is used to determine the optimal values ​​for the autoregressive order, the difference order, and the moving average order. Specifically, stationarity tests are performed sequentially starting from the first difference until they pass, ultimately yielding a stationary sequence.

[0021] Finally, the optimal values ​​for the autoregression order, the difference order, and the moving average order are determined based on the minimum mean square error criterion.

[0022] An improved particle swarm optimization algorithm is employed, using the charging and discharging power of electric vehicles as the position value of each particle. The calculation dimension is the number of electric vehicles per unit area multiplied by a set number of particles. By iterating through randomly generated particle values, the optimal position p traversed by each particle at time t is determined. best and the best position g in the group best .

[0023] The method for updating the particle's velocity and position is as follows:

[0024] v i,j (t+1)=v i,j (t)+c1·rand()·[p i,j -x i,j [(t)]+c2·rand()·[p g,j -x i,j (t)]

[0025] x i,j (t+1)=x i,j (t)+v i,j (t+1)

[0026] v i.j ∈[v min ,v max ]

[0027] x i,j ∈X

[0028] Where c1 and c2 are the first and second positive acceleration constants, respectively, and rand() is a randomly generated particle value; v i,j (t) represents the velocity of the particle with coordinates (i, j) at time t. i,j (t+1) represents the velocity of the particle at coordinate (i, j) at time t+1, p i,jLet p be the position of the particle with coordinates (i, j). g,j x is the optimal position of the j-th particle in the swarm. i,j (t) represents the particle value at coordinates (i, j), v min v is the minimum particle velocity. max X represents the maximum particle velocity, and X is the set of possible particle positions.

[0029] The constraints required for updating particle velocity and position include electric vehicle parameter constraints, electric vehicle user demand constraints, charging and discharging power constraints of charging piles, and distribution transformer capacity constraints.

[0030] The daytime charging and discharging plan of electric vehicles in the region generated in step 3 and the basic load of the regional power grid collected in step 2 are superimposed to form the daytime forecast total load of the regional power grid. The dispatch plan is formulated based on the daytime forecast total load of the regional power grid and passes the safety verification.

[0031] Security checks include N-1 security checks, which are generally performed using a two-stage robust optimization model. The objective function is:

[0032]

[0033] Where, N f T is the set of all power supply nodes in the power system. f This is the set of time periods during which the system malfunctioned. The load loss of node n in the system when it fails at time t; the target of the safety check is the total load loss of the entire system during the failure period. Keep it within the set range.

[0034] In step 5, the daily charge and discharge plan for electric vehicles is rolled over in 15-minute increments.

[0035] The model predictive control method based on a controllable autoregressive integral moving average model is used to optimize the charging and discharging power of the clustered electric vehicles in the current rolling time domain, using the following discrete difference equation:

[0036] A(z -1 y(t)=B(z) -1 u(t-1)+C(z) -1 )ω(t) / Δ

[0037]

[0038]

[0039]

[0040] In the formula, A(z) -1 B(z) -1 ), C(z) -1 ) respectively represent the contents of operator z -1 The first, second, and third polynomials, N a N b N c ω(t) represents the order of the first polynomial, the second polynomial, and the third polynomial, respectively; u(t) and y(t) represent the input and output of the controlled object, respectively. During the rolling optimization process, u(t) is the daytime charging and discharging plan of the electric vehicle in the region at time t, y(t) is the real-time charging and discharging plan of the electric vehicle at time t, and ω(t) is the noise vector; a ii b jj c xx Let ii represent the coefficient sequences of the first polynomial, the second polynomial, and the third polynomial, respectively, where ii = 1, 2, ..., N. a jj=0,1…N b ,xx=1,2…N c z is a random operator; through continuous difference iteration, a convergent solution is eventually obtained.

[0041] A power system dispatching framework system incorporating electric vehicle resources includes a historical information collection module, a day-ahead forecasting module for charging and discharging quantities, a module for setting control targets and control requirements, a module for setting day-ahead charging and discharging plans for electric vehicles, a module for setting day-ahead dispatching plans for the power system, a safety verification module, and a rolling correction module for electric vehicle charging and discharging plans.

[0042] The historical data collected by the historical information collection module includes location data, average travel mileage, start charging time, estimated departure time, start SOC, and demand SOC. The collected data is then input into the charge / discharge day-ahead prediction module.

[0043] The day-ahead prediction module for charge and discharge capacity calculates and predicts the charge and discharge capacity of electric vehicles and inputs the predicted data into the control target and control demand formulation module.

[0044] The control target and control demand formulation module obtains the control demand of the regional power grid, calculates the demand capacity, and inputs the results into the electric vehicle daytime charging and discharging plan formulation module.

[0045] The electric vehicle day-ahead charging and discharging plan formulation module adopts an improved particle swarm optimization algorithm, which uses the charging and discharging power of electric vehicles as the position value of each particle, and the calculation dimension is the number of electric vehicles multiplied by the set number of particles; and the calculation results are input into the power system day-ahead dispatch plan formulation module.

[0046] The power system day-ahead dispatch planning module overlays the generated electric vehicle day-ahead charging and discharging plan with the collected regional power grid base load to form the regional power grid day-ahead forecast total load, and inputs the total load to the safety verification module;

[0047] After the safety verification module performs a safety verification on the total load using the N-1 safety verification method, it inputs the verification result into the rolling correction module of the electric vehicle charging and discharging plan.

[0048] The rolling correction module for electric vehicle charging and discharging plans, based on the day-ahead charging and discharging plans of electric vehicles and the day-ahead scheduling plans of the power system, combined with the charging and discharging information of electric vehicles in advance and in real time, performs rolling corrections of the electric vehicle charging and discharging plans in 15-minute increments to achieve real-time scheduling.

[0049] The beneficial effects of this invention are that, compared with the prior art, this invention:

[0050] 1. This invention proposes a power system dispatching framework method that incorporates electric vehicle resources. By integrating electric vehicles as source-side resources into the original dispatching process at the dispatching level, the distributed energy storage characteristics of large-scale electric vehicles can be better utilized to coordinate the regulation of grid load, provide auxiliary services to the grid, and improve the regional grid regulation capability.

[0051] 2. The power system dispatching framework method incorporating electric vehicle resources of the present invention can integrate electric vehicle resources into the power system dispatching plan formulation process, which is conducive to optimizing the regional power grid's capacity to accept electric vehicle loads and resources, and provides support for the stable operation of the regional power system. Attached Figure Description

[0052] Figure 1 A flowchart of a power system dispatching framework method incorporating electric vehicle resources according to the present invention;

[0053] Figure 2 This invention provides a method for predicting the charging and discharging load and behavior of electric vehicles based on the ARIMA algorithm in one embodiment.

[0054] Figure 3 The output curves of the main distribution transformers within the region and the optimized output curves are shown in one embodiment of the present invention. Detailed Implementation

[0055] The present application will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention, and should not be construed as limiting the scope of protection of the present application.

[0056] like Figure 1 A power system dispatch framework method incorporating electric vehicle resources includes the following steps:

[0057] Step 1: Based on the historical charging and discharging information of electric vehicles collected from each electric vehicle charging station in the region, make a day-ahead forecast of the charging and discharging volume of each electric vehicle charging station in the region, using the charging station or electric vehicle load aggregator as the unit.

[0058] Furthermore, the content and indicators for collecting electric vehicle charging and discharging information in step 1 are as follows, conforming to GB / T32960 "Technical Specifications for Electric Vehicle Remote Service and Management System":

[0059] Table 1. Content and Indicators of Electric Vehicle Charging and Discharging Information Collection

[0060]

[0061] Based on the collected historical charging and discharging information of electric vehicles, a corresponding time series table can be generated, dividing the day into 96 points with a step size of 15 minutes. An example table format is as follows:

[0062] Table 2. Example of historical charging and discharging information for electric vehicles (taking charging power as an example).

[0063]

[0064]

[0065] Using stations or electric vehicle load aggregators as units, a time-series forecasting method is adopted to predict the day-ahead charging and discharging of electric vehicles.

[0066] Preferably, a time-series prediction method for charge and discharge capacity based on differential integration can be used to generate day-ahead prediction results for the charge and discharge capacity of each charging station in the region based on historical data of electric vehicle charge and discharge over a continuous time period.

[0067] Preferably, information on electric vehicles at each electric vehicle charging station in the region should first be collected according to Table 1, especially information on electric vehicle charging and discharging loads and charging behavior with strong regularity, to obtain time-series historical information. This allows for the use of predictive models to predict electric vehicle charging and discharging information. Specific steps are as follows: Figure 2 As shown.

[0068] Those skilled in the art can select a prediction model based on the actual situation. The embodiment provided is only a preferred solution and should not be construed as a necessary limitation on the scope of protection of this invention.

[0069] First, the time series of historical information on electric vehicles is smoothed using a difference algorithm. When there are obvious outliers in the data series, causing irregular peaks or troughs, difference processing can smooth out these outliers and reduce their impact on the prediction series. The prediction model can be denoted as ARIMA(p,d,q), which uses a difference-integrated moving average autoregressive algorithm, where p, d, and q are the autoregression order, difference order, and moving average order, respectively. Preferably, different orders should be used depending on the charging and discharging information of electric vehicles; in this paper, p=3, d=2, and q=1 are used. The original information time series is denoted as Ori={T}. i}(i=1,2,…,N), where N is the number of electric vehicles per unit area.

[0070] The general form of ARIMA(p,d,q) is:

[0071] φ(B)(1-B) d t i =θ(B)ε t

[0072]

[0073] θ(B)=1-θ1B-θ2B 2 -…-θ q B q

[0074] Where B is the shift operator, ε t To obey N(0,σ) 2 Error term of normal distribution; θ i Represents the autocorrelation coefficient of the corresponding time series, i = 1, 2, ..., q; φ(B) represents the moving average coefficient of the corresponding time series, j = 1, 2, ..., p; φ(B) represents the autoregressive model of the time series, and θ(B) represents the moving average model of the time series.

[0075] Secondly, the stabilization effect of the difference equation requires the application of the Enhanced Dickey-Fowler test (ADF) to obtain a rough range of the predicted model order. Then, the minimum information criterion is used to determine the optimal model order. Specifically, stabilization tests are performed sequentially starting from the first-order difference until a stationary sequence is obtained.

[0076] Finally, the optimal model parameters are determined based on the minimum mean square error criterion, and the prediction results of electric vehicle charging and discharging quantity and charging and discharging behavior are reconstructed by the inverse operation of the difference.

[0077] Step 2: Determine the regional power grid control objectives and control requirements based on the regional power grid's basic load information;

[0078] The regional power grid regulation targets and regulation needs include peak shaving periods and peak shaving demand capacity, respectively.

[0079] Furthermore, in step 2, it is necessary to calculate various operating indicators of the regional power grid, and use the judgment criteria formed by the expert evaluation method to obtain the regional power grid's control demand, and calculate the demand capacity. The judgment criteria are shown in the table below:

[0080] Table 3. Criteria for Judging the Regulation and Control Needs of Regional Power Grids

[0081]

[0082] Step 3: Based on the control requirements determined in Step 2, control is carried out on the basis of the electric vehicle charging and discharging situation predicted in Step 1 to form the daily charging and discharging plan for electric vehicles;

[0083] Furthermore, in step 3, adjustments are made based on the predicted charging and discharging situation of electric vehicles to form a day-ahead charging and discharging plan for electric vehicles, and different electric vehicle charging and discharging plan optimization methods are adopted for different grid control objectives.

[0084] Preferably, different electric vehicle charging and discharging plan optimization methods are adopted for different power grid control objectives. Taking peak shaving as an example, a day-ahead charging and discharging plan generation method based on an improved particle swarm optimization (PSO) algorithm can be used. Those skilled in the art can also choose the algorithm for generating the charging and discharging plan according to the actual situation. The embodiment given here is only a preferred solution and should not be construed as a necessary limitation on the scope of protection of this invention.

[0085] When using the improved particle swarm optimization algorithm, the charging and discharging power of the electric vehicles can be used as the position value of each particle, and the calculation dimension is the number of electric vehicles multiplied by the set number of particles. By iterating over the randomly generated particle values, the optimal position p traversed by each particle at time t is determined. best and the best position g in the group best The velocity and position of each particle are updated according to the following rules.

[0086] v i,j (t+1)=v i,j (t)+c1·rand()·[p i,j -x i,j [(t)]+c2·rand()·[p g,j -x i,j (t)]

[0087] x i,j (t+1)=x i,j (t)+v i,j (t+1)

[0088] v i.j ∈[v min ,v max ]

[0089] x i,j ∈X

[0090] Where c1 and c2 are the first and second positive acceleration constants, respectively, and rand() is a randomly generated particle value; v i,j (t) represents the velocity of the particle with coordinates (i, j) at time t. i,j (t+1) represents the velocity of the particle at coordinate (i, j) at time t+1, p i,j Let p be the position of the particle with coordinates (i, j). g,j x is the optimal position of the j-th particle in the swarm. i,j (t) represents the particle value at coordinates (i, j), v min v is the minimum particle velocity. max X represents the maximum particle velocity, and X is the set of possible particle positions.

[0091] Ultimately, a day-ahead plan for electric vehicles to participate in grid auxiliary peak shaving is generated, namely, a 96-point charging and discharging power plan for electric vehicles. At the same time, some constraints also need to be considered, including electric vehicle parameter constraints, electric vehicle user demand constraints, charging and discharging power constraints of charging piles, and regional distribution transformer capacity constraints.

[0092] Step 4: Combining the regional power grid basic load information in Step 2 and the regional electric vehicle daytime charging and discharging plan generated in Step 3, a daytime dispatch plan for the power system is formed, and the safety and control effect are verified and evaluated.

[0093] Preferably, the daytime charging and discharging plan of electric vehicles generated in step 3 and the basic load of the regional power grid collected in step 1 should be superimposed to form the daytime predicted total load of the regional power grid. The dispatch plan is formulated according to this predicted total load. Those skilled in the art can set the dispatch plan according to the actual situation. The dispatch plan given in this invention is detailed in the embodiments given in the invention book. After that, a security check is performed, which includes N-1 security checks, etc. Generally, a two-stage robust optimization model is used for N-1 security checks, and the objective function is:

[0094]

[0095] Where, N f T is the set of all power supply nodes in the power system. f This is the set of time periods during which the system malfunctioned. This represents the load loss of node n in the system when it fails at time t. The objective of the safety check is the total load loss of the entire system during the failure period. It can be controlled within a certain range.

[0096] Step 5: Based on the daytime charging and discharging plan of electric vehicles and the daytime dispatching plan of the power system in Steps 3 and 4, and combined with the charging and discharging information of electric vehicles in advance and in real time, the charging and discharging plan of electric vehicles is rolled over in 15-minute increments to finally achieve real-time dispatching.

[0097] Furthermore, the rolling correction of the electric vehicle charging and discharging plan in step 5 is characterized by real-time correction of the electric vehicle charging and discharging plan under the determined power system dispatch plan to meet the operational requirements of the power system. Based on the information uploaded by electric vehicles at public charging stations, residential charging locations, and dedicated charging stations, and combined with the pre-assessment results of the day-ahead dispatchability of electric vehicles, model predictive control technology based on the controllable autoregressive integral moving average (CARIMA) model is used to optimize the charging and discharging power of the cluster of electric vehicles in the current rolling time domain, and the control command for the first time period of the cluster of electric vehicles is issued. At the next moment, combining the state information of newly arrived electric vehicles and the updated state information of electric vehicles in the grid, the state of the cluster of electric vehicles in the rolling time domain is predicted and its optimal charging and discharging power is solved, and the control command for the first time period of the electric vehicle is issued. This process is repeated cyclically to continuously update the electric vehicle state information and simultaneously issue new optimal commands, thereby achieving rolling optimization of the electric vehicle dispatching scheme. The rolling time domain can be set to 15 minutes. The mathematical model of CARIMA optimization can be described by discrete difference equations:

[0098] A(z -1 y(t)=B(z) -1 u(t-1)+C(z) -1 )ω(t) / Δ

[0099]

[0100]

[0101]

[0102] Where A(z) -1 B(z) -1 ), C(z) -1 ) respectively represent the contents of operator z -1 The first, second, and third polynomials, N a N b N cω is the corresponding order; u(t) and y(t) are the input and output of the controlled object, respectively. In the rolling optimization process, u(t) is the daytime charging and discharging plan of the electric vehicle at time t, y(t) is the real-time charging and discharging plan of the electric vehicle at time t, and ω(t) is the noise vector. ii b jj c xx Let ii represent the coefficient sequences of the first polynomial, the second polynomial, and the third polynomial, respectively, where ii = 1, 2, ..., N. a jj=0,1…N b ,xx=1,2…N c z is a random operator; through continuous difference iteration, a convergent solution is eventually obtained.

[0103] Example:

[0104] The process of generating a power grid dispatch plan in a certain area of ​​Nanjing, Jiangsu Province, was simulated. The selected area includes 1232 electric vehicles. Figure 3 The dashed line represents the original dispatch curve for this region without considering electric vehicle resources. The peak load exceeds the total regional distribution capacity, requiring measures such as increasing inter-regional interconnection power, adjusting local energy storage, and increasing renewable energy output to meet the demand. This demonstrates that the primary demand for regional power grid regulation is peak shaving. After optimization using the method described in this patent, the optimized output curves of the main distribution transformers in the region are as follows: Figure 3 As shown by the solid line, it can be seen that the optimization method presented in this paper will reduce the power supply load, improve the power regulation margin, and help optimize the regional power grid's ability to accommodate electric vehicle loads and resources, thus providing support for the stable operation of the regional power system.

[0105] This invention also discloses a power system dispatching system including an electric vehicle resource dispatching module. New functional modules are added to the original power dispatching system, including an electric vehicle historical information collection module, a day-ahead charging and discharging quantity forecasting module, a regional power grid control target and control demand formulation module, an electric vehicle day-ahead charging and discharging plan formulation module, and an electric vehicle charging and discharging plan rolling correction module. The original power system day-ahead dispatching plan formulation module and safety verification module are improved.

[0106] The historical data collected by the historical information collection module includes location data, average travel mileage, start charging time, estimated departure time, start SOC, and demand SOC. The collected data is then input into the charge / discharge day-ahead prediction module.

[0107] The day-ahead prediction module for charge and discharge capacity calculates and predicts the charge and discharge capacity of electric vehicles and inputs the predicted data into the control target and control demand formulation module.

[0108] The control target and control demand formulation module obtains the control demand of the regional power grid, calculates the demand capacity, and inputs the results into the electric vehicle daytime charging and discharging plan formulation module.

[0109] The electric vehicle day-ahead charging and discharging plan formulation module adopts an improved particle swarm optimization algorithm, which uses the charging and discharging power of electric vehicles as the position value of each particle, and the calculation dimension is the number of electric vehicles multiplied by the set number of particles; and the calculation results are input into the power system day-ahead dispatch plan formulation module.

[0110] The power system day-ahead dispatch planning module overlays the generated electric vehicle day-ahead charging and discharging plan with the collected regional power grid base load to form the regional power grid day-ahead forecast total load, and inputs the total load to the safety verification module;

[0111] After the safety verification module performs a safety verification on the total load using the N-1 safety verification method, it inputs the verification result into the rolling correction module of the electric vehicle charging and discharging plan.

[0112] After the day-ahead dispatch plan of the power system is formed, it is necessary to continue to collect the charging and discharging information of electric vehicles at the real-time level. The rolling correction module of the electric vehicle charging and discharging plan, based on the day-ahead charging and discharging plan of electric vehicles and the day-ahead dispatch plan of the power system, combined with the charging and discharging information of electric vehicles in the previous time and in real time, performs rolling correction of the electric vehicle charging and discharging plan in 15-minute increments to achieve real-time dispatch.

[0113] The applicant of this invention has provided a detailed description of the embodiments of the invention in conjunction with the accompanying drawings. However, those skilled in the art should understand that the above embodiments are merely preferred embodiments of the invention. The detailed description is only intended to help readers better understand the spirit of the invention and is not intended to limit the scope of protection of the invention. On the contrary, any improvements or modifications made based on the inventive spirit of the invention should fall within the scope of protection of the invention.

Claims

1. A power system dispatching framework method incorporating electric vehicle resources, characterized in that, The power system dispatch framework method includes: Step 1: Obtain time-series historical information based on the charging and discharging historical information of electric vehicles collected from each electric vehicle charging station in the region, and make day-ahead forecasts of the charging and discharging volume of each electric vehicle charging station in the region, using electric vehicle charging stations or electric vehicle load aggregators as units. The forecasting model used in the current forecast is ,in , , These are the autoregression order, differencing order, and moving average order, respectively; the historical information of the time series is denoted as... , This represents the time-series historical information of the i-th electric vehicle; The number of electric vehicles per unit area; Specifically: In the formula, For the shift operator, To obey The error term of the normal distribution; Represents the autocorrelation coefficient of the corresponding time series, i=1,2,…q; Represents the moving average coefficient of the corresponding time series, j=1,2…p; This represents the autoregressive model for the time series. This represents the moving average model for the time series; Step 2: Determine the regional power grid control objectives and control requirements based on the regional power grid's basic load information; Step 3: Based on the control requirements determined in Step 2, control the charging and discharging volume of each electric vehicle charging station in the region obtained from the day-ahead forecast in Step 1 to form a day-ahead charging and discharging plan for electric vehicles in the region. Step 4: Combining the regional power grid basic load information from Step 2 and the regional electric vehicle daytime charging and discharging plan generated in Step 3, a daytime dispatch plan for the power system is formed, and its safety is verified. Step 5: Based on the daytime charging and discharging plan of electric vehicles in the region and the daytime dispatch plan of the power system, collect and combine the charging and discharging information of electric vehicles in advance and in real time, and perform rolling correction of the daytime charging and discharging plan of electric vehicles to achieve real-time dispatch.

2. The power system dispatching framework method including electric vehicle resources according to claim 1, characterized in that, In step 1, the collected historical charging and discharging information of electric vehicles includes location data, average travel mileage, start charging time, estimated departure time, start SOC, and demand SOC.

3. The power system dispatching framework method including electric vehicle resources according to claim 1, characterized in that, In step 1, the time series historical information of electric vehicle charging and discharging history is stabilized using a differential algorithm.

4. The power system dispatching framework method including electric vehicle resources according to claim 1, characterized in that, The enhanced Dickey-Fowler test is used to determine the range of values ​​for the autoregressive order, the difference order, and the moving average order in the prediction model. Then, the minimum information criterion is used to determine the optimal values ​​for the autoregressive order, the difference order, and the moving average order. Specifically, stationarity tests are performed sequentially starting from the first difference until they pass, ultimately yielding a stationary sequence. Finally, the optimal values ​​for the autoregression order, the difference order, and the moving average order are determined based on the minimum mean square error criterion.

5. The power system dispatching framework method including electric vehicle resources according to claim 4, characterized in that, An improved particle swarm optimization algorithm is employed, using the charging and discharging power of electric vehicles as the position value of each particle. The calculation dimension is the number of electric vehicles per unit area multiplied by a set number of particles. By iterating through randomly generated particle values, the optimal position traversed by each particle at time t is determined. and the best position in the group .

6. The power system dispatching framework method including electric vehicle resources according to claim 5, characterized in that, The method for updating the velocity and position of the particles is as follows: in, and These are the first positive acceleration constant and the second positive acceleration constant, respectively. These are randomly generated particle values; The coordinates at time t are ( i , j The velocity of the particles, For the first t The coordinates at time +1 are ( i , j The particle velocity, The coordinates are ( i , j The position of the particle. For the first in the group j The optimal position of each particle The coordinates are ( i , j The particle value of ) This represents the minimum particle velocity. This represents the maximum particle velocity. It is the set of possible positions of a particle.

7. A power system dispatching framework method incorporating electric vehicle resources according to claim 6, characterized in that, The constraints required for the method of updating the velocity and position of the particles include electric vehicle parameter constraints, electric vehicle user demand constraints, charging and discharging power constraints of the charging piles, and distribution transformer capacity constraints.

8. A power system dispatching framework method incorporating electric vehicle resources according to claim 1 or 7, characterized in that, The daytime charging and discharging plan of electric vehicles in the region generated in step 3 and the basic load of the regional power grid collected in step 2 are superimposed to form the daytime forecast total load of the regional power grid. The dispatch plan is formulated based on the daytime forecast total load of the regional power grid and passes the safety verification.

9. A power system dispatching framework method incorporating electric vehicle resources according to claim 8, characterized in that, The security verification includes Security checks are generally performed using a two-stage robust optimization model. Security verification, the objective function is: in, It is the set of all power supply nodes in the power system. This is the set of time periods during which the system malfunctioned. For nodes in the system In time The amount of load loss during a failure; the goal of safety verification is the total amount of load loss of the entire system during a failure. Keep it within the set range.

10. A power system dispatching framework method incorporating electric vehicle resources according to claim 1, characterized in that, In step 5, the daily charge and discharge plan for electric vehicles is rolled over in 15-minute increments.

11. A power system dispatching framework method incorporating electric vehicle resources according to claim 1 or 10, characterized in that, The model predictive control method based on a controllable autoregressive integral moving average model is used to optimize the charging and discharging power of the clustered electric vehicles in the current rolling time domain, using the following discrete difference equation: In the formula, , , They respectively represent the presence of operators The first, second and third polynomials, , , These are the orders of the first polynomial, the second polynomial, and the third polynomial, respectively. and These are the input and output of the controlled object, respectively. During the scrolling optimization process... For the day-ahead charging and discharging schedule of electric vehicles in the region at time t, The real-time charging and discharging plan for the electric vehicle at time t. This is the noise vector; , , Let ii represent the coefficient sequences of the first polynomial, the second polynomial, and the third polynomial, respectively, where ii = 1, 2, ..., N. a jj=0,1…N b , xx=1,2…N c ; It is a random operator; through continuous difference iteration, a convergent solution is eventually obtained.

12. A power system dispatching framework system incorporating electric vehicle resources, implemented based on the power system dispatching framework method incorporating electric vehicle resources as described in any one of claims 1-11, characterized in that, The power system dispatch framework includes a historical information collection module, a day-ahead forecasting module for charging and discharging quantities, a module for setting control targets and control requirements, a module for setting day-ahead charging and discharging plans for electric vehicles, a module for setting day-ahead dispatching plans for the power system, a safety verification module, and a rolling correction module for electric vehicle charging and discharging plans. The historical data collected by the historical information collection module includes location data, average travel mileage, start charging time, estimated departure time, start SOC, and demand SOC. The collected data is then input into the charge / discharge day-ahead prediction module. The day-ahead prediction module for charge and discharge capacity calculates and predicts the charge and discharge capacity of electric vehicles and inputs the predicted data into the control target and control demand formulation module. The control target and control demand formulation module obtains the control demand of the regional power grid, calculates the demand capacity, and inputs the results into the electric vehicle daytime charging and discharging plan formulation module. The electric vehicle day-ahead charging and discharging plan formulation module adopts an improved particle swarm optimization algorithm, which uses the charging and discharging power of electric vehicles as the position value of each particle, and the calculation dimension is the number of electric vehicles multiplied by the set number of particles; and the calculation results are input into the power system day-ahead dispatch plan formulation module. The power system day-ahead dispatch planning module overlays the generated electric vehicle day-ahead charging and discharging plan with the collected regional power grid base load to form the regional power grid day-ahead forecast total load, and inputs the total load to the safety verification module; After the safety verification module performs a safety verification on the total load using the N-1 safety verification method, it inputs the verification result into the rolling correction module of the electric vehicle charging and discharging plan. The rolling correction module for electric vehicle charging and discharging plans, based on the day-ahead charging and discharging plans of electric vehicles and the day-ahead scheduling plans of the power system, combined with the charging and discharging information of electric vehicles in advance and in real time, performs rolling corrections of the electric vehicle charging and discharging plans in 15-minute increments to achieve real-time scheduling.

Images