Double-layer control method, system and equipment for charging of electric vehicle in active power distribution network and storage medium

Abstract

The invention discloses a double-layer control method, system and device for electric vehicle charging in an active power distribution network and a storage medium. The double-layer control method comprises the following steps: predicting distributed energy power; the initial capacity and the initial SOC of the battery of the electric vehicle are obtained; performing long-time scale optimization control according to the predicted power data of the distributed energy and the initial capacity and SOC data of the battery of the electric vehicle to obtain a basic charging scheme of the electric vehicle; obtaining a power prediction error according to the real-time power and the prediction power of the distributed energy, and extracting a high-frequency component of the power error by adopting wavelet decomposition; and performing short-time scale optimal control by using the error high-frequency component and the real-time SOC of the battery of the electric vehicle to obtain an adjustment scheme of charging of the electric vehicle so as to obtain a final charging scheme of the electric vehicle. Reasonable charging and discharging control of the electric vehicle under the power distribution network can be achieved, EV charging participates in ADN control, DER integration can be enhanced, and ADN operation is effectively improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of demand-side response control in distribution networks, and in particular to a double-layer control method, system, device and storage medium for charging electric vehicles in an active distribution network based on wavelet decomposition. Background Art

[0002] Active distribution networks (ADNs) are suitable for integrating distributed energy resources (DERs) and multi-type loads due to their flexible topology and abundant controllable resources. The inherent variability and uncertainty of DERs pose challenges to the generation-load balance of ADNs. Additional power reserves from other controllable resources are needed to compensate for the power fluctuations of DERs. At the same time, electric vehicles (EVs) are developing rapidly in ADNs. With the large-scale application of electric vehicles and the rapid development of clean energy from distributed photovoltaic power generation, building electric vehicle charging stations with distributed photovoltaic power generation as the main energy source can solve the problems of clean energy consumption and the impact of electric vehicle charging on the power grid. Since EVs are usually charged simultaneously in time and space, EV charging is a pulse load, which will also bring significant pressure to ADNs if not properly controlled. On the one hand, electric vehicles adopt an unordered charging method of charging as soon as they are used up. When the charging time coincides with the peak period of the distribution network, the upper limit of the peak will be increased, causing an impact on the distribution network; on the other hand, since the charging piles of electric vehicles contain a large number of nonlinear components, a large amount of harmonic current will be generated during charging, affecting the power quality. In addition, the frequent changes in charging power not only aggravate the power imbalance, but also shorten the service life of EV batteries.

[0003] Therefore, it is necessary to develop an EV charging strategy to promote the integration of distributed energy and electric vehicles and reduce the adverse effects of distributed energy and electric vehicles on the operation of active distribution networks. Summary of the invention

[0004] The technical problem to be solved by the present invention is to overcome the defects of the above-mentioned prior art and provide a two-layer control method, system, device and storage medium for charging electric vehicles in an active distribution network based on wavelet decomposition. The method adopts a two-layer electric vehicle charging strategy, that is, based on the predicted distributed energy power, a multi-objective optimization model that weighs the net power fluctuation and the change of the electric vehicle charging power is established in the outer layer to obtain the basic charging plan of the electric vehicle on a large time scale; in the inner layer, wavelet decomposition is used to extract the high-frequency power prediction error of the distributed energy, and the basic charging plan of the electric vehicle is adjusted according to the extracted high-frequency component and the real-time charge state of the electric vehicle battery to obtain the final charging plan of the electric vehicle.

[0005] In a first aspect, the present invention provides a double-layer control method for charging an electric vehicle in an active power distribution network, comprising:

[0006] Predicting distributed energy power;

[0007] Obtain the initial capacity and initial state of charge of the electric vehicle battery;

[0008] According to the predicted power of distributed energy and the initial capacity and initial state of charge of electric vehicle batteries, long-term optimization control is carried out to obtain the basic charging plan of electric vehicles on an hourly time scale;

[0009] The power prediction error on a minute-level time scale is obtained based on the real-time power and predicted power of distributed energy, and the high-frequency component of the power prediction error is extracted using wavelet decomposition.

[0010] Utilizing the high-frequency component of the power prediction error and the real-time state of charge of the electric vehicle battery, short-time-scale optimal control is performed to obtain an adjustment plan for charging the electric vehicle;

[0011] The basic charging scheme for electric vehicles is combined with the adjustment scheme for charging electric vehicles to obtain the final charging scheme for electric vehicles.

[0012] Furthermore, the long-term optimization control adopts a multi-objective optimization model to minimize the fluctuation of the predicted power of distributed energy and the change of the charging power of electric vehicles. The objective function is the net power variation coefficient C n and the electric vehicle charging power variation coefficient C e , respectively represent the smoothness of the power fluctuation of distributed energy prediction and the limit of the change of electric vehicle charging power. The main purpose of the multi-objective optimization model is to optimize the charging power of electric vehicles, and the decision variable is the charging power of electric vehicles.

[0013] The net power variation coefficient and the electric vehicle charging power variation coefficient reflect the fluctuations of the net power and the electric vehicle charging power respectively. Taking both into account in the objective function can obtain a basic charging scheme that can both smooth the predicted power fluctuations of distributed energy and minimize the impact of electric vehicle charging.

[0014] Furthermore, the net power is calculated from the predicted power of distributed energy resources and the optimized charging / discharging power of electric vehicles, which is expressed as:

[0015] P n (t) = P d (t)-P e (t)

[0016] Among them, P n (t), P d (t) and P e (t) are the net power at time t, the predicted power of distributed energy resources and the optimized charging / discharging power of electric vehicles, in kW;

[0017] Net power variation coefficient C n Calculated as:

[0018]

[0019] Among them, u n is the average value of net power, in kW; T is the optimization time length, i.e. 12 hours;

[0020] Electric vehicle charging / discharging power variation coefficient C e The calculation method is as follows:

[0021]

[0022] Among them, u e It is the average value of the charging / discharging power of the electric vehicle, in kW.

[0023] Furthermore, the objective function of the multi-objective optimization model is expressed as:

[0024] minf o (P e,t )=min(f1,f2)=min(C n ,C e )

[0025] The constraints are expressed as:

[0026] P dis,max ≤P e (t)≤P ch,max

[0027] SOC min ≤SOC(t)≤SOC max

[0028]

[0029] Among them, P ch,max and P dis,max are the maximum charging power and the maximum discharging power, respectively, in kW; SOC(t) is the equivalent SOC at time t; SOC min and SOC max are the minimum allowable SOC and the maximum allowable SOC respectively; E r is the rated capacity of the connected electric vehicle battery, in kW·h; E(t-1) is the energy stored in the electric vehicle battery at time t-1, in kW·h; η ch and η dis They are the charging efficiency and discharging efficiency of electric vehicles respectively.

[0030] Furthermore, an improved MOPSO algorithm is used to solve the multi-objective optimization model. The elite archive in the MOPSO algorithm is used to store the optimal Pareto solution set during the iteration process, and the speed and position of the particles are updated according to the local and global Pareto solutions. In order to improve the fixed inertia weight coefficient and learning factor in the MOPSO algorithm that causes the optimization algorithm to fall into the local optimum, a dynamic inertia factor is used to adjust the weight value to balance the global and local, thereby improving the algorithm's search speed and local optimization capabilities.

[0031] The dynamic inertia factor is used to adjust the weight value, so that the MOPSO algorithm can jump out of the local optimal position with a higher probability when it falls into the local optimal solution, ensuring that the algorithm has good search ability during the optimization process.

[0032] Furthermore, based on the improved MOPSO algorithm, the Pareto solution set is obtained; the decision factor of the Pareto optimal solution is calculated to select the electric vehicle charging scheme, which is expressed as:

[0033]

[0034]

[0035] d f,i =a1f g,1,i +a2f g,2,i

[0036] a1+a2=1

[0037] Among them, f g,1,i and f g,2,i is the generalized fitness value of the i-th optimal Pareto solution; f 1,i and f 2,i is the fitness value of the i-th Pareto optimal solution; f 1,max and f 2,max are the maximum fitness values ​​in the Pareto solution set; f 1,min and f 2,min are the minimum fitness values ​​in the Pareto solution set; a1 and a2 are weight coefficients; d f,i is the decision factor of the i-th Pareto optimal solution, and the solution with the smallest decision factor is selected as the basic charging solution for electric vehicles.

[0038] Furthermore, the goal of wavelet decomposition is to extract the high-frequency components of the power prediction error of distributed energy. Assuming ψ(t)∈L 2 (R) is a mother wavelet,

[0039]

[0040] Among them, ψ(ω) is the Fourier transform of ψ(t), then, relative to the mother wavelet ψ(t), the discrete wavelet representation of the sequence s(t) is:

[0041]

[0042] Among them, a and b are the scale factor and the translation factor respectively; by adjusting the scale factor a and the translation factor b, the sequence s(t) is decomposed and the corresponding change characteristics are obtained at a specific frequency;

[0043] The high-frequency subsequence e of the power prediction error is extracted using a specific scaling factor a and a translation factor b. h (t), and is updated as the data window slides; then, the last high-frequency subsequence e h (t l ) is used to l Adjust the charging sequence of electric vehicles;

[0044] Based on the real-time SOC of the electric vehicle battery, the adjusted power is calculated. h (t l,i )≥0,

[0045]

[0046] When e h (t l,i )<0,

[0047]

[0048] Among them, t l,i is the time of the last prediction error of the ith subsequence; ΔP(t l,i ) is the time t l,i Adjusted power in kW; SOC r and SOC o They are real-time SOC and long-time scale optimized SOC; d m is the maximum allowable deviation of SOC;

[0049] The charging power of electric vehicles is adjusted to:

[0050] P C (t l,i )=P E (t l,i )+ΔP(t l,i )

[0051] Among them, P C (t l,i ) is the time t l,i The final charging power; P E (tl,i ) is the time t l,i Long time scale optimized charging power.

[0052] In a second aspect, the present invention provides a two-layer control system for charging electric vehicles in an active power distribution network, comprising:

[0053] Power prediction unit: used to predict distributed energy power;

[0054] Initial parameter acquisition unit: used to obtain the initial capacity and initial state of charge of the electric vehicle battery;

[0055] Basic charging plan acquisition unit: Based on the predicted power of distributed energy and the initial capacity and initial charge state of the electric vehicle battery, long-term optimization control is performed to obtain the basic charging plan for the electric vehicle;

[0056] High-frequency component extraction unit: obtain the power prediction error of distributed energy according to the real-time power of distributed energy and the predicted power of distributed energy, and use wavelet decomposition to extract the high-frequency component of the power prediction error;

[0057] A charging adjustment scheme acquisition unit: using the high-frequency component of the power prediction error and the real-time state of charge of the electric vehicle battery, performs short-time optimal control to obtain an adjustment scheme for charging the electric vehicle;

[0058] Final charging plan acquisition unit: The basic charging plan for the electric vehicle is combined with the adjustment plan for charging the electric vehicle to obtain the final charging plan for the electric vehicle.

[0059] In a third aspect, the present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method when executing the computer program.

[0060] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the described method when executed by a processor.

[0061] The beneficial effects of the present invention are as follows:

[0062] 1) Suppress the fluctuation of distributed energy prediction power and the impact of electric vehicle charging power in the distribution network: construct the net power variation coefficient and the electric vehicle charging power variation coefficient, take both into account in the outer objective function, and obtain the basic charging plan of electric vehicles based on the predicted power of distributed energy. The inner structure adjusts the charging plan of electric vehicles based on the power prediction error of distributed energy.

[0063] 2) Correcting the charging power of electric vehicles: Decomposing the distributed energy power prediction error based on the wavelet decomposition method, adjusting the electric vehicle power based on the high-frequency component and considering the real-time SOC of the electric vehicle battery;

[0064] 3) Improve the search performance and convergence performance of the algorithm: In order to improve the fixed inertia weight coefficient and learning factor that causes the optimization algorithm to fall into the local optimum, a dynamic inertia factor is used to adjust the weight value, balance the global and local, and improve the algorithm's search speed and local optimization capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0066] Figure 1 It is a logic block diagram of a double-layer control method for charging electric vehicles in an active power distribution network of the present invention;

[0067] Figure 2 It is a logic block diagram of the improved MOPSO algorithm of the present invention;

[0068] Figure 3 It is a diagram of the double-layer control system for charging electric vehicles in an active power distribution network of the present invention;

[0069] Figure 4 Schematic diagram of the Pareto frontier in an application example of the present invention;

[0070] Figure 5 It is a schematic diagram of decision factors of the optimal Pareto solution in the application example of the present invention;

[0071] Figure 6 is a schematic diagram of a basic EV charging scheme in an application example of the present invention;

[0072] Figure 7 is a schematic diagram of net power curves of different control methods in an application example of the present invention;

[0073] Figure 8 A schematic diagram of a logical structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0074] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0075] Example 1

[0076] This embodiment is a two-layer control method for charging electric vehicles in an active distribution network. Figure 1 As shown, it adopts double-layer control.

[0077] The steps of the outer control are as follows: predict the power of distributed energy; obtain the initial capacity and initial SOC of the electric vehicle battery; perform long-term optimization control based on the predicted power data of distributed energy and the initial capacity and SOC data of the electric vehicle battery to obtain the basic charging plan for electric vehicles on an hourly time scale.

[0078] The steps of the inner layer control are as follows: obtaining the power prediction error on a minute time scale according to the real-time power and predicted power of distributed energy, and extracting the high-frequency component of the power prediction error by using wavelet decomposition; performing short-time-scale optimal control by using the high-frequency component of the power prediction error and the real-time SOC of the electric vehicle battery, and obtaining an adjustment plan for the charging of the electric vehicle;

[0079] The basic charging scheme for electric vehicles is combined with the adjustment scheme for charging electric vehicles to obtain the final charging scheme for electric vehicles.

[0080] The long-term optimization control described above adopts a multi-objective optimization model to minimize the fluctuation of the predicted power of distributed energy resources and the change of the charging power of electric vehicles. The objective function is the net power variation coefficient C n and the electric vehicle charging power variation coefficient C e , respectively represent the smoothness of distributed energy power fluctuations and the limit of electric vehicle charging power changes. The net power is calculated by the predicted power of distributed energy and the optimized charging / discharging power of electric vehicles, expressed as:

[0081] P n (t) = P d (t)-P e (t)

[0082] Among them, P n (t), P d (t) and P e(t) are the net power at time t, the predicted power of distributed energy resources and the optimized charging / discharging power of electric vehicles, in kW.

[0083] Net power variation coefficient C n Calculated as:

[0084]

[0085] Among them, u n is the average value of net power in kW; T is the optimization time length, i.e. 12 hours.

[0086] Electric vehicle charging / discharging power variation coefficient C e The calculation method is as follows:

[0087]

[0088] Among them, u e It is the average value of the charging / discharging power of the electric vehicle, in kW.

[0089] The objective function of the multi-objective optimization model is expressed as:

[0090] minf o (P e,t )=min(f1,f2)=min(C n ,C e )

[0091] The constraints are expressed as:

[0092] P dis,max ≤P e (t)≤P ch,max

[0093] SOC min ≤SOC(t)≤SOC max

[0094]

[0095] Among them, P ch,max and P dis,max are the maximum charging power and the maximum discharging power, respectively, in kW; SOC(t) is the equivalent SOC at time t; SOC min and SOC max are the minimum allowable SOC and the maximum allowable SOC respectively; E r is the rated capacity of the connected electric vehicle battery, in kW·h; E(t-1) is the energy stored in the electric vehicle battery at time t-1, in kW·h; η ch and η disThey are the charging efficiency and discharging efficiency of electric vehicles respectively.

[0096] To solve the above optimization model, the improved MOPSO algorithm is used, and the elite archive is used to store the optimal Pareto solution set during the iteration process; in addition, the speed and position of the particle are updated according to the local and global Pareto solutions, which can be expressed as:

[0097]

[0098] Among them, v i k and v i k+1 are the velocities of particle i in the kth and k+1th iterations respectively; x i k and x i k+1 are the positions of particle i in the kth and k+1th iterations respectively; x g k is the global Pareto solution in the kth iteration; x l,i k is the local Pareto solution of particle i in the kth iteration; w k is the inertia weight in the kth iteration; c1 and c2 are the global and local acceleration coefficients respectively; r1 and r2 are random numbers uniformly distributed in [0,1].

[0099] In order to improve the fixed inertia weight coefficient and learning factor, the optimization algorithm falls into the local optimum, making the algorithm less accurate. This embodiment uses a dynamic inertia factor to adjust the weight value, balance the global and local, and improve the algorithm's search speed and local optimization capabilities. The inertia weight w is adjusted as the number of iterations k increases, which can be expressed as:

[0100]

[0101] Among them, w k+1 is the inertia weight in the kth iteration; w max is the maximum inertia weight, w min is the minimum inertia weight; N max is the total number of iterations.

[0102] Adjusting the learning factor makes particles pay attention to the population position information, which can better maintain the convergence speed and search ability, and adjust the fitness of the group, which can be expressed as:

[0103]

[0104] The larger the c1, the wider the search range of the particles; the larger the c2, the closer the particles converge to the global optimum; max and c min are the upper and lower limits of the learning factor respectively.

[0105] The calculation logic of the improved MOPSO algorithm is as follows: Figure 2 As shown, the specific steps are as follows:

[0106] 1. Input raw data and information: initial capacity and SOC of electric vehicles, predicted power of distributed energy, etc.

[0107] 2. Initialize the particle swarm. Initializing the speed and position of all particles is equivalent to obtaining all charging schemes for electric vehicles.

[0108] 3. Initialize the elite file. Calculate the individual extreme value of the particle and the global extreme value of the population, and store them.

[0109] 4. Update the inertia weight coefficient and learning factor. Use dynamic inertia factors to adjust the weights and learning factors.

[0110] 5. Update the particle's position and velocity.

[0111] 6. Extreme value update. Filter out the optimal Pareto solution that meets the objective function and constraints, and update the historical optimum and global optimum.

[0112] 7. Determine the maximum number of iterations. Determine whether the maximum number of iterations is met. If not, continue to iterate.

[0113] 8. The final solution is determined. Stop when the maximum number of loop iterations is met, and output the optimal solution and objective function value.

[0114] Based on the MOPSO algorithm, the Pareto solution set is obtained to determine the basic charging scheme for electric vehicles on a 1-hour time scale; the decision factor is calculated to select the charging scheme for electric vehicles, which is expressed as:

[0115]

[0116] d f,i =a1f g,1,i +a2f g,2,i

[0117] a1+a2=1

[0118] Among them, f g,1,i and f g,2,i is the generalized fitness value of the i-th optimal Pareto solution; f 1,i and f 2,i is the fitness value of the i-th Pareto optimal solution; f 1,max and f 2,max are the maximum fitness values ​​in the Pareto solution set; f 1,min and f 2,minare the minimum fitness values ​​in the Pareto solution set; a1 and a2 are weight coefficients; d f,i is the decision factor of the i-th Pareto optimal solution, and the solution with the smallest decision factor is selected as the basic charging solution for electric vehicles.

[0119] The goal of wavelet decomposition is to extract the high-frequency component of the power prediction error of distributed energy. The prediction error is the power prediction error of distributed energy on a 1-minute time scale. Assume that ψ(t)∈L 2 (R) is a mother wavelet,

[0120]

[0121] Among them, ψ(ω) is the Fourier transform of ψ(t), then, relative to the mother wavelet ψ(t), the discrete wavelet representation of the sequence s(t) is:

[0122]

[0123] Among them, a and b are the scale factor and translation factor respectively; by adjusting the scale factor a and the translation factor b, the sequence s(t) is decomposed and the corresponding change characteristics are obtained at a specific frequency; in order to extract the high-frequency prediction error e(t), a 15-minute sliding data window is used. Therefore, the high-frequency subsequence e of the power prediction error is extracted using the specific scale factor a and the translation factor b. h (t), and is updated as the data window slides; then, the last high-frequency subsequence e h (t l ) is used to l Adjust the charging sequence of electric vehicles.

[0124] Taking into account the real-time SOC of the EV battery, the adjusted power is calculated. h (t l,i )≥0,

[0125]

[0126] When e h (t l,i )<0,

[0127]

[0128] Among them, t l,i is the time of the last power prediction error of the ith subsequence; ΔP(t l,i ) is the time t l,i Adjusted power in kW; SOC r and SOC o They are real-time SOC and long-time scale optimized SOC; dm is the maximum allowable deviation of SOC.

[0129] Therefore, the charging power of the electric vehicle is adjusted to:

[0130] P C (t l,i )=P E (t l,i )+ΔP(t l,i )

[0131] Among them, P C (t l,i ) is the time t l,i The final charging power; P E (t l,i ) is the time t l,i Long time scale optimized charging power.

[0132] Example 2

[0133] This embodiment is a two-layer control system for charging electric vehicles in an active power distribution network. Figure 3 As shown, it consists of a power prediction unit, an initial parameter acquisition unit, a basic charging plan acquisition unit, a high-frequency component extraction unit, a charging adjustment plan acquisition unit and a final charging plan acquisition unit.

[0134] Power prediction unit: used to predict distributed energy power;

[0135] Initial parameter acquisition unit: used to obtain the initial capacity and initial SOC of the electric vehicle battery;

[0136] Basic charging plan acquisition unit: Based on the predicted power data of distributed energy and the initial capacity and SOC data of electric vehicle batteries, long-term optimization control is performed to obtain the basic charging plan for electric vehicles;

[0137] High-frequency component extraction unit: obtain the power prediction error of distributed energy according to the real-time power of distributed energy and the predicted power of distributed energy, and use wavelet decomposition to extract the high-frequency component of the power prediction error;

[0138] A charging adjustment scheme acquisition unit: using the high-frequency component of the power prediction error and the real-time SOC of the electric vehicle battery, performs short-time-scale optimal control to obtain an adjustment scheme for charging the electric vehicle;

[0139] Final charging plan acquisition unit: The basic charging plan for the electric vehicle is combined with the adjustment plan for charging the electric vehicle to obtain the final charging plan for the electric vehicle.

[0140] In the basic charging scheme acquisition unit, the long-term optimization control adopts a multi-objective optimization model to minimize the predicted power fluctuation of distributed energy and the charging power variation of electric vehicles. The objective function is the net power variation coefficient C n and the electric vehicle charging power variation coefficient C e , respectively represent the smoothness of distributed energy power fluctuations and the limit of electric vehicle charging power changes. The main purpose of the multi-objective optimization model is to optimize the charging power of electric vehicles, and the decision variable is the charging power of electric vehicles.

[0141] Specifically, the net power is calculated by the predicted power of distributed energy and the optimized charging / discharging power of electric vehicles, which is expressed as:

[0142] P n (t) = P d (t)-P e (t)

[0143] Among them, P n (t), P d (t) and P e (t) are the net power at time t, the predicted power of distributed energy resources and the optimized charging / discharging power of electric vehicles, in kW;

[0144] Net power variation coefficient C n Calculated as:

[0145]

[0146] Among them, u n is the average value of net power, in kW; T is the optimization time length, i.e. 12 hours;

[0147] Electric vehicle charging / discharging power variation coefficient C e The calculation method is as follows:

[0148]

[0149] Among them, u e It is the average value of the charging / discharging power of the electric vehicle, in kW.

[0150] Specifically, the objective function of the multi-objective optimization model is expressed as:

[0151] minf o (P e,t )=min(f1,f2)=min(C n ,C e )

[0152] The constraints are expressed as:

[0153] P dis,max ≤P e (t)≤P ch,max

[0154] SOC min ≤SOC(t)≤SOC max

[0155]

[0156] Among them, P ch,max and P dis,max are the maximum charging power and the maximum discharging power, respectively, in kW; SOC(t) is the equivalent SOC at time t; SOC min and SOC max are the minimum allowable SOC and the maximum allowable SOC respectively; E r is the rated capacity of the connected electric vehicle battery, in kW·h; E(t-1) is the energy stored in the electric vehicle battery at time t-1, in kW·h; η ch and η dis They are the charging efficiency and discharging efficiency of electric vehicles respectively.

[0157] Specifically, an improved MOPSO algorithm is used to solve the multi-objective optimization model. The elite archive is used to store the optimal Pareto solution set during the iteration process, and the velocity and position of the particle are updated according to the local and global Pareto solutions. In order to improve the fixed inertia weight coefficient and learning factor that causes the optimization algorithm to fall into the local optimum, a dynamic inertia factor is used to adjust the weight value to balance the global and local, thereby improving the algorithm's search speed and local optimization ability.

[0158] The dynamic inertia factor is used to adjust the weight value, so that the MOPSO algorithm can jump out of the local optimal position with a higher probability when it falls into the local optimal solution, ensuring that the algorithm has good search ability during the optimization process.

[0159] Specifically, based on the improved MOPSO algorithm, the Pareto solution set is obtained; the decision factor is calculated to select the electric vehicle charging scheme, which is expressed as:

[0160]

[0161] d f,i =a1f g,1,i +a2f g,2,i

[0162] a1+a2=1

[0163] Among them, f g,1,i and f g,2,i is the generalized fitness value of the i-th optimal Pareto solution; f1,i and f 2,i is the fitness value of the i-th Pareto optimal solution; f 1,max and f 2,max are the maximum fitness values ​​in the Pareto solution set; f 1,min and f 2,min are the minimum fitness values ​​in the Pareto solution set; a1 and a2 are weight coefficients; d f,i is the decision factor of the i-th Pareto optimal solution, and the solution with the smallest decision factor is selected as the basic charging solution for electric vehicles.

[0164] In the high-frequency component extraction unit, the goal of wavelet decomposition is to improve the high-frequency components of distributed energy power prediction errors. Assuming ψ(t)∈L 2 (R) is a mother wavelet,

[0165]

[0166] Among them, ψ(ω) is the Fourier transform of ψ(t), then, relative to the mother wavelet ψ(t), the discrete wavelet representation of the sequence s(t) is:

[0167]

[0168] Among them, a and b are the scale factor and translation factor respectively; by adjusting the scale factor a and the translation factor b, the sequence s(t) is decomposed and the corresponding change characteristics are obtained at a specific frequency.

[0169] In the charging adjustment scheme acquisition unit, a specific proportional factor a and a translation factor b are used to extract a high-frequency subsequence e of the power prediction error. h (t), and is updated as the data window slides; then, the last high-frequency subsequence e h (t l ) is used to l Adjust the charging sequence of electric vehicles;

[0170] Taking into account the real-time SOC of the EV battery, the adjusted power is calculated. h (t l,i )≥0,

[0171]

[0172] When e h (t l,i )<0,

[0173]

[0174] Among them, t l,iis the time of the last power prediction error of the ith subsequence; ΔP(t l,i ) is the time t l,i Adjusted power in kW; SOC r and SOC o They are real-time SOC and long-time scale optimized SOC; d m is the maximum allowable deviation of SOC.

[0175] In the final charging scheme acquisition unit, the charging power of the electric vehicle is adjusted to:

[0176] P C (t l,i )=P E (t l,i )+ΔP(t l,i )

[0177] Among them, P C (t l,i ) is the time t l,i The final charging power; P E (t l,i ) is the time t l,i Long time scale optimized charging power.

[0178] Application Examples

[0179] The double-layer control method described in Example 1 is used as follows.

[0180] Based on the predicted power data of distributed energy, long-term optimization control is performed to optimize the charging power of electric vehicle batteries. In the outer control, the battery capacity and SOC of the electric vehicle at 19:00 are collected, and the distributed energy power of each hour from 19:00 to 7:00 is predicted. In addition, a multi-objective optimization model is established to minimize the DER power fluctuation and EV charging power variation. The decision variable is the EV charging / discharging power P for 1 hour from 19:00 to 7:00. E The objective function is the net power variation coefficient C n and the electric vehicle charging / discharging power variation coefficient C e , represent the smoothness of DER power fluctuations and the limitation of EV power variation, respectively.

[0181] like Figure 4 and 5 As shown in Figure 2, the decision factors of the Pareto frontier and the optimal Pareto solution obtained by the improved MOPSO algorithm.

[0182] In the optimization of the outer control, the maximum number of iterations N of the improved MOPSO is m Set the Pixels to 500 and the Particles to 70.max and c min The values ​​are 2 and 0.2 respectively, and the maximum weight w max and the minimum inertia weight w min are set to 30 and 0 respectively. The weight coefficients a1 and a2 are set to 0.5. The Pareto frontier obtained is as follows Figure 4 As shown. The decision factor d of the optimal Pareto solution f like Figure 5 shown.

[0183] like Figure 6 The following is a basic scheme for charging electric vehicles.

[0184] When the net power C n and EV charging power C e When the coefficients of variation are 0.5254 and 0.3080 respectively, the smallest decision factor is 0.4669, and this solution is selected as the basic charging solution.

[0185] like Figure 7 Shown are the net power curves using the method of the present invention and the outer layer control method, as well as the case without EV charging control.

[0186] The above basic charging scheme is adjusted considering high frequency components and real-time SOC. Figure 7 The net power curves using the method of the present invention and the outer control method, as well as the case without EV charging control, are shown in Figure 2. From the power curves, it is obvious that the power fluctuation is the smallest using the method of the present invention. More specifically, the standard deviations of the above three scenarios are 706.60, 722.93, and 723.66, respectively.

[0187] Example 3

[0188] This embodiment provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor in communication. The memory stores a computer program executable by the at least one processor, and when the computer program is executed by the at least one processor, the electronic device executes the method according to Embodiment 1 of the present invention.

[0189] Example 4

[0190] This embodiment provides a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor of a computer, is used to cause the computer to execute the method according to Embodiment 1 of the present invention.

[0191] refer to Figure 8, a block diagram of an electronic device 400 that can be used as a server or client of the present invention will now be described, which is an example of a hardware device that can be applied to various aspects of the present invention. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0192] like Figure 8 As shown, the electronic device 400 includes a computing unit 401, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 402 or a computer program loaded from a storage unit 408 into a random access memory (RAM) 403. In the RAM 403, various programs and data required for the operation of the electronic device 400 can also be stored. The computing unit 401, the ROM 402, and the RAM 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0193] A plurality of components in the electronic device 400 are connected to the I / O interface 405, including: an input unit 406, an output unit 407, a storage unit 408, and a communication unit 409. The input unit 406 may be any type of device capable of inputting information to the electronic device 400, and the input unit 406 may receive input digital or character information, and generate key signal inputs related to user settings and / or function control of the electronic device. The output unit 407 may be any type of device capable of presenting information, and may include, but is not limited to, a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 408 may include, but is not limited to, a disk, an optical disk. The communication unit 409 allows the electronic device 400 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks, and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication transceiver, and / or a chipset, such as a Bluetooth™ device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0194] The computing unit 401 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 401 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 401 performs the various methods and processes described above. For example, in some embodiments, the aforementioned metal corrosion prediction method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 408. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 400 via the ROM 402 and / or the communication unit 409. In some embodiments, the computing unit 401 may be configured to perform the aforementioned metal corrosion prediction method in any other appropriate manner (e.g., by means of firmware).

[0195] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0196] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0197] As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and / or device (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0198] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0199] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0200] A computer system may include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other.

[0201] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A two-tier control method for charging electric vehicles in an active distribution network, characterized in that: include: Predicting distributed energy power; Obtain the initial capacity and initial SOC of the electric vehicle battery; According to the predicted power of distributed energy and the initial capacity and initial SOC of electric vehicle batteries, long-term optimization control is carried out to obtain the basic charging plan of electric vehicles on an hourly time scale; The power prediction error on a minute-level time scale is obtained based on the real-time power and predicted power of distributed energy, and the high-frequency component of the power prediction error is extracted using wavelet decomposition. Utilizing the high-frequency component of the power prediction error and the real-time SOC of the electric vehicle battery, short-time-scale optimal control is performed to obtain an adjustment plan for charging the electric vehicle; The basic charging scheme for electric vehicles is combined with the adjustment scheme for charging electric vehicles to obtain the final charging scheme for electric vehicles.

2. The double-layer control method for charging electric vehicles in an active power distribution network according to claim 1, characterized in that: The long-term optimization control described above adopts a multi-objective optimization model to minimize the fluctuation of the predicted power of distributed energy resources and the change of the charging power of electric vehicles. The objective function is the net power variation coefficient C n and the electric vehicle charging power variation coefficient C e , representing the smoothness of the predicted power fluctuation of distributed energy resources and the limit of the charging power variation of electric vehicles, respectively.

3. The double-layer control method for charging electric vehicles in an active power distribution network according to claim 2, characterized in that: The net power is calculated from the predicted power of distributed energy and the optimized charging / discharging power of electric vehicles, and is expressed as: P n (t)=P d (t)-P e (t) Among them, P n (t), P d (t) and P e (t) are the net power at time t, the predicted power of distributed energy resources and the optimized charging / discharging power of electric vehicles, in kW; Net power variation coefficient C n Calculated as: Among them, u n is the average value of net power, in kW; T is the optimization time length, i.e. 12 hours; Electric vehicle charging / discharging power variation coefficient C e The calculation method is as follows: Among them, u e It is the average value of the charging / discharging power of the electric vehicle, in kW.

4. The double-layer control method for charging electric vehicles in an active power distribution network according to claim 3, characterized in that: The objective function of the multi-objective optimization model is expressed as: myf o (P e,t )=min(f1,f2)=min(C n ,C e ) The constraints are expressed as: P dis,max ≤P e (t)≤P ch,max SOC min ≤SOC(t)≤SOC max Among them, P ch,max and P dis,max are the maximum charging power and the maximum discharging power, respectively, in kW; SOC(t) is the equivalent SOC at time t; SOC min and SOC max are the minimum allowable SOC and the maximum allowable SOC respectively; E r is the rated capacity of the connected electric vehicle battery, in kW·h; E(t-1) is the energy stored in the electric vehicle battery at time t-1, in kW·h; η ch and η dis They are the charging efficiency and discharging efficiency of electric vehicles respectively.

5. The double-layer control method for charging electric vehicles in an active power distribution network according to claim 4, characterized in that: The MOPSO algorithm is used to solve the multi-objective optimization model. The elite archive in the MOPSO algorithm is used to store the optimal Pareto solution set during the iteration process, and the speed and position of the particles are updated according to the local and global Pareto solutions. In order to improve the fixed inertia weight coefficient and learning factor in the MOPSO algorithm that causes the optimization algorithm to fall into the local optimum, the dynamic inertia factor is used to adjust the weight value to balance the global and local.

6. The double-layer control method for charging electric vehicles in an active power distribution network according to claim 5, characterized in that: Based on the MOPSO algorithm, the Pareto solution set is obtained; the decision factor of the Pareto optimal solution is calculated to select the electric vehicle charging scheme, which is expressed as: <h2 style=";text-align:left;direction:ltr">d<h2 style=";text-align:left;direction:ltr"> f,i <h2 style=";text-align:left;direction:ltr"> =a1f<h2 style=";text-align:left;direction:ltr"> g,1,i <h2 style=";text-align:left;direction:ltr"> +a2f<h2 style=";text-align:left;direction:ltr"> g,2,i a1+a2=1 Among them, f g,1,i and f g,2,i is the generalized fitness value of the i-th optimal Pareto solution; f 1,i and f 2,i is the fitness value of the i-th Pareto optimal solution; f 1,max and f 2,max are the maximum fitness values ​​in the Pareto solution set; f 1,min and f 2,min are the minimum fitness values ​​in the Pareto solution set; a1 and a2 are weight coefficients; d f,i is the decision factor of the i-th Pareto optimal solution, and the solution with the smallest decision factor is selected as the basic charging solution for electric vehicles.

7. The double-layer control method for charging electric vehicles in an active power distribution network according to claim 1, characterized in that: The goal of wavelet decomposition is to extract the high-frequency components of the power prediction error of distributed energy resources. Assume that ψ(t)∈L 2 (R) is a mother wavelet, Among them, ψ(ω) is the Fourier transform of ψ(t), then, relative to the mother wavelet ψ(t), the discrete wavelet representation of the sequence s(t) is: Among them, a and b are the scale factor and the translation factor respectively; by adjusting the scale factor a and the translation factor b, the sequence s(t) is decomposed and the corresponding change characteristics are obtained at a specific frequency; The high-frequency subsequence e of the power prediction error is extracted using a specific scaling factor a and a translation factor b. h (t), and is updated as the data window slides; then, the last high-frequency subsequence e h (t l ) is used to l Adjust the charging sequence of electric vehicles; Based on the real-time SOC of the electric vehicle battery, the adjusted power is calculated. h (t l,i )≥0, When e h (t l,i )<0, Among them, t l,i is the time of the last power prediction error of the ith subsequence; ΔP(t l,i ) is the time t l,i Adjusted power in kW; SOC r and SOC o They are real-time SOC and long-time scale optimized SOC; d m is the maximum allowable deviation of SOC; The charging power of electric vehicles is adjusted to: P C (t l,i )=P E (t l,i )+ΔP(t l,i ) Among them, P C (t l,i ) is the time t l,i The final charging power; P E (t l,i ) is the time t l,i Long time scale optimized charging power.

8. A two-layer control system for charging electric vehicles in an active power distribution network, used to implement the method described in any one of claims 1 to 8, characterized in that: include: Power prediction unit: used to predict distributed energy power; Initial parameter acquisition unit: used to obtain the initial capacity and initial SOC of the electric vehicle battery; Basic charging plan acquisition unit: Based on the predicted power of distributed energy and the initial capacity and initial SOC of the electric vehicle battery, long-term optimization control is performed to obtain the basic charging plan for the electric vehicle; High-frequency component extraction unit: obtain the power prediction error of distributed energy according to the real-time power of distributed energy and the predicted power of distributed energy, and use wavelet decomposition to extract the high-frequency component of the power prediction error; A charging adjustment scheme acquisition unit: using the high-frequency component of the power prediction error and the real-time SOC of the electric vehicle battery, performs short-time-scale optimal control to obtain an adjustment scheme for charging the electric vehicle; Final charging plan acquisition unit: The basic charging plan for the electric vehicle is combined with the adjustment plan for charging the electric vehicle to obtain the final charging plan for the electric vehicle.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images