An optimized scheduling method for guiding the orderly charging of electric vehicles

Abstract

An optimized scheduling method for guiding the orderly charging of electric vehicles includes the following steps: S1. Establish an unordered charging load prediction model for electric vehicles based on the Monte Carlo method; S2. Introduce a carbon trading mechanism into the unordered charging load prediction model of electric vehicles in step S1 to calculate the carbon emissions and carbon quotas of power generation enterprises; S3. Introduce a reward coefficient ξ1 and a carbon trading cost growth rate τ1 into the conventional stepped carbon price to establish a reward and punishment stepped carbon price model; S4. Establish a dynamic time-of-use electricity price model based on the reward and punishment stepped carbon price; S5. Establish an electricity price regulation model for electric vehicles; S6. Establish a multi-objective optimization function for the orderly charging of electric vehicles; S7. Adopt a multi-objective non-dominated genetic NSGA-II solution algorithm; The present invention guides electric vehicle users to perform orderly charging behaviors by using a regulation method based on the reward and punishment stepped carbon price and the dynamic time-of-use electricity price, thereby effectively achieving the effect of peak shaving and valley filling of the power grid and reducing the fluctuation of the power grid.

Description

Technical Field

[0001] The present invention relates to the technical field of charging scheduling, and in particular to an optimized scheduling method for guiding the orderly charging of electric vehicles. Background Art

[0002] In recent years, electric vehicles have received extensive attention. Among them, pure electric vehicles, as representatives of new energy vehicles, replace oil with electricity compared with traditional fuel vehicles and have many advantages, such as low driving costs, less greenhouse gas emissions, low noise, high energy conversion rate, and diversified energy sources. Therefore, electric vehicles have become an important means to achieve the decarbonization path in the power and transportation fields. As an important part of China's energy Internet, the future ownership of electric vehicles will continue to increase. However, the disorderly charging of a large number of electric vehicles will lead to an increase in the peak load of the power system, affect the power quality, and pose new challenges to the vehicle-grid scheduling of the distribution network. For example, during the peak electricity consumption period at night, the charging of electric vehicles has led to the phenomenon of "peak on peak", causing a greater impact on the power grid. How to enable electric vehicles and the distribution network to achieve safe, reliable, and economic positive interaction has become an urgent problem to be solved at present.

[0003] According to consumer psychology, price differences will affect users' consumption patterns. According to price changes, users choose to accept the current price or change the charging time, etc. Traditional research focuses on using time-of-use electricity prices to regulate the charging behavior of electric vehicle users, thereby guiding the orderly charging of electric vehicles. However, the current situation of power generation enterprises in the upstream power supply of electric vehicles is still the production of fossil energy combustion, and a large amount of CO 2 is generated. At present, the domestic power industry mainly adopts the method of free allocation for the initial carbon emission quota allocation. When the carbon emissions generated by power generation enterprises during power generation exceed the carbon quota, they need to purchase carbon emission rights, and the abundant carbon emission quotas can be traded and profited in the carbon market. Therefore, it is crucial to study the orderly charging scheduling mechanism of the new power market - carbon market under the carbon trading mechanism with electric vehicles as the main body.

[0004] Chinese Patent (Publication No.: CN2022103462134) discloses an optimized scheduling method for a virtual power plant considering demand response and carbon trading. This patent proposes an optimized scheduling method for a virtual power plant in which flexible and controllable loads such as electric vehicles and water-cooled air conditioners and distributed power sources participate in the power market and the carbon trading market. However, the above patent only considers that the carbon emission trading price adopts a fixed value, so the constraint or incentive effect on the carbon emissions of enterprises is not reflected.

[0005] Chinese Patent (Publication No.: CN2018110881615) discloses a method for constructing a comprehensive energy system model considering the carbon trading mechanism and demand response; this patent proposes a method for constructing the uncertainty of the multi-energy loads of electricity, heat, and cold on the demand side using the information gap theory, and establishing a scheduling model for the comprehensive energy system with stepped carbon trading and demand response; however, the above patent only considers the correlation between the stepped carbon trading price and the actual carbon emissions of the comprehensive energy system, without considering the internal coupling relationship between the carbon trading volume price and the time-of-use electricity price, as well as the quantitative analysis of the key parameters in the carbon trading mechanism. Summary of the Invention

[0006] In order to overcome the deficiencies in the background technology, the present invention discloses an optimized scheduling method for guiding the orderly charging of electric vehicles.

[0007] To achieve the above-mentioned invention purpose, the present invention adopts the following technical solutions:

[0008] An optimized scheduling method for guiding the orderly charging of electric vehicles, comprising the following steps:

[0009] S1. Establish an unordered charging load prediction model for electric vehicles based on the Monte Carlo method;

[0010] S2. Introduce the carbon trading mechanism into the unordered charging load prediction model of electric vehicles in step S1, and calculate the carbon emissions and carbon quotas of power generation enterprises;

[0011] E ct = δ P P Mt

[0012] E Pt = c 1 P Mt

[0013] Where: E ct is the carbon quota obtained at time t; E Pt is the carbon emissions at time t; P Mt is the active power output of the generator set at time t, with the unit of kW, δ P is the carbon emission quota per unit of power generation, taking 0.798 kg / kWh; c 1 is the power carbon emission factor, representing the CO 2 emission per degree of electricity, taking 0.8386 kg / kWh;

[0014] S3. Introduce the reward coefficient ξ 1 and the carbon trading cost growth rate τ 1 into the conventional stepped carbon price, and establish a reward and punishment stepped carbon price model;

[0015] S4. Establish a dynamic time-of-use electricity price model based on the reward and punishment stepped carbon price;

[0016]

[0017] X t0 = X 0 λ 1 / λ 0

[0018] a 1 = a 0 τ 1 / τ 0

[0019] b 1 = b 0 ξ 1 / ξ 0

[0020]

[0021] P max = max(P t0 + P tEV )

[0022] P min = min(P t0 + P tEV )

[0023] Where: X 0 is the initial electricity price; Y 0 is the electricity price decline interval, in MW, obtained by dividing the difference between the average daily total load and the minimum value by the number of increase intervals; Y 1 is the electricity price increase interval, obtained by dividing the difference between the maximum daily total load and the average value by the number of increase intervals, in MW; P av represents the average value of the daily total load; P t0 is the original grid load without electric vehicle load at time t, in MW; P tEV is the electric vehicle load at time t, in MW; T is the number of segments of each day; X t0 represents time t, the initial electricity price under the influence of the reward and punishment ladder carbon price; P max represents the maximum total load within a day, in MW; P min represents the minimum total load within a day, in MW; λ 0 is the initial carbon trading cost base price; τ 0 is the initial carbon trading cost growth rate, taking 0.3; a 1 represents the electricity price growth coefficient under the influence of the reward and punishment ladder carbon price; a 0 is the initial electricity price growth coefficient; ξ 1 is the initial reward coefficient, taking 0.2; b 1Indicates the electricity price decline coefficient under the influence of the reward and punishment ladder carbon price; b 0 Is the initial electricity price decline coefficient;

[0024] S5. Establish an electric vehicle electricity price regulation model;

[0025]

[0026] P tEVnew = ρP tEV

[0027] P 总 = P t0 + P tEVnew

[0028] Among them, P tEVnew Is the electric vehicle load power at time t after the electricity price changes; P 总 Is the total grid load;

[0029] S6. Establish a multi-objective optimization function for the orderly charging of electric vehicles;

[0030]

[0031] Among them: Is the average value of the total grid load after time-of-use electricity price scheduling, with the unit of kW; T is the number of segments of each day's time; F 1 Is the objective function with the minimum peak-valley mean square deviation of the superposition of the original grid load and the electric vehicle charging load; F 2 Is the objective function with the minimum total cost of electric vehicle charging users, with the unit of ten thousand yuan; F 3 Is the objective function with the minimum carbon trading cost of power generation enterprises, with the unit of ten thousand yuan;

[0032] S7. Adopt the multi-objective non-dominated genetic NSGA-II solution algorithm; obtain the objective functions F 1 ~ F 3 through steps S1 to S6; loop the above steps N times to obtain the objective functions of N populations; perform non-dominated sorting on the objective functions of N populations and calculate the crowding degree of population individuals; then enter the population evolution loop stage, evaluate the objective functions of the offspring, and obtain the three objective functions of the offspring; merge the parent population (N) and the offspring population (N) (2N), perform non-dominated sorting and crowding degree calculation on the three objective functions of the merged population again, and obtain a new parent population (N) through the elite retention strategy until the evolution generation reaches the set maximum generation; after the iteration ends, merge the Pareto solutions obtained in each run, apply non-dominated sorting to the merged Pareto solutions, and screen out the Pareto optimal solutions with rank 1.

[0033] Preferably, in step S1, the return time t of residential users per day is set 0 obeys a normal distribution with an expectation of μ t and a variance of σ t 2 t ~ N(μ t , σ t 2 ). Taking μ t = 17.6 and σ t 2 = 3.4, the probability density function of the electric vehicle return time t 0 is:

[0034]

[0035] The daily driving mileage S of electric vehicle users obeys a lognormal distribution with an expectation of μ s and a variance of σ s 2 S ~ log(μ s , σ s 2 ). Taking μ s = 3.2 and σ s 2 = 0.88, the probability density function of the daily driving mileage S of electric vehicles is:

[0036]

[0037] The charging duration required for an electric vehicle with a daily driving mileage S is:

[0038]

[0039] where: E is the power consumption per kilometer of the electric vehicle, in kWh / km; P c is the charging power of the electric vehicle, in kW; η c is the charging efficiency of the electric vehicle;

[0040] According to the Monte Carlo method, the return time and daily driving mileage of a single electric vehicle are simulated, and the daily charging load of the electric vehicle is:

[0041]

[0042] where: P ci is the charging power of the i-th electric vehicle, in kW, and I i is used to determine whether the i-th vehicle is charging at time t. I i = 0 indicates not charging, and I i = 1 indicates charging.

[0043] Preferably, the state of charge (SOC) constraint of the electric vehicle in step S1 is:

[0044] SOC min ≤SOC≤SOC max

[0045] where: SOC min and SOC max are the minimum and maximum values of the remaining power of the electric vehicle, respectively, with the unit of Ah.

[0046] Preferably, the active power output of the generator set in step S2 satisfies the power balance constraint of P Mt =P t0 +P tEV .

[0047] Preferably, the reward and punishment ladder carbon price model in step S3 is:

[0048]

[0049] E t =E Pt -E ct

[0050] where: C t CO2 is the carbon trading cost at time t, with the unit of yuan, λ 1 is the base price of the carbon trading cost, with the unit of yuan / ton; E t is the difference between the total carbon emissions and the free-allocated carbon quota at time t, with the unit of ton; τ 1 is the growth rate of the carbon trading cost for each stage; ξ 1 is the reward coefficient, L 0 is the length of the negative interval of the carbon cost, with the unit of ton; L 1 is the length of the positive interval of the carbon cost, with the unit of ton.

[0051] Preferably, the electricity price constraint in step S5 is:

[0052] X min ≤X t ≤X max

[0053] where: X min and X max are the marginal cost electricity price and the maximum electricity price, respectively, with the unit of yuan.

[0054] Preferably, in step S7, first, the initial electricity price is set through the value of the initial carbon trading cost base price λ 0 , and the initial electricity price growth coefficient a 0 and the initial electricity price decline coefficient b 0The upper and lower limits are used to achieve the purpose of electricity price constraints, and at the same time, the population size N and the original grid load P without the charging load of electric vehicles are initialized. t0 。

[0055] Due to the above-described technical solution, the present invention has the following beneficial effects:

[0056] An optimized scheduling method for guiding the orderly charging of electric vehicles disclosed by the present invention can effectively guide electric vehicle users to perform orderly charging behaviors by using a regulation method based on a reward and punishment-based stepped carbon price and a dynamic time-of-use electricity price, thereby effectively achieving the effect of peak shaving and valley filling of the power grid and reducing the fluctuation of the power grid; using the change of the reward and punishment-based stepped carbon price and the load to regulate the electricity price, and finally realizing the regulation of the charging behaviors of electric vehicle users; respectively taking the minimum total load mean square error, the minimum total user charging cost, and the minimum carbon trading cost as the objective functions, and using the NSGA-II multi-objective optimization genetic algorithm for solution; compared with the traditional NSGA algorithm, the computational complexity of the NSGA-II algorithm is greatly reduced, so that the population individuals in the quasi-Pareto domain can be evenly extended to the entire Pareto domain, avoiding the situation where the result falls into a local optimum and ensuring the diversity of the population. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a flowchart of the prediction of the unordered charging load of electric vehicles based on the Monte Carlo method;

[0058] Figure 2 is a prediction graph of the charging load under the unordered charging of electric vehicles;

[0059] Figure 3 is the relationship between the carbon trading price and the carbon trading volume;

[0060] Figure 4 is the electricity price response curve of electric vehicles;

[0061] Figure 5 is a flowchart of the method for the orderly charging scheduling of electric vehicles based on the stepped carbon price and the time-of-use electricity price;

[0062] Figure 6 is a flowchart of the three-objective NSGA-II solution algorithm;

[0063] Figure 7 is the original load of the power grid;

[0064] Figure 8 is the comparison of the optimization of the total load scheduling of the power grid under three scenarios in the first embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0065] The present invention can be explained in detail through the following embodiments. The purpose of disclosing the present invention is to protect all technical improvements within the scope of the present invention. In the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "front", "rear", "left", "right", etc. indicating the orientation or positional relationship, it is only corresponding to the drawings of the present application for the convenience of describing the present invention, rather than indicating or implying that the device or element referred to must have a specific orientation.

[0066] An optimized scheduling method for guiding the orderly charging of electric vehicles, comprising the following steps:

[0067] S1. Establish an unordered charging load prediction model for electric vehicles based on the Monte Carlo method; set the return time t of daily residential users 0 obeying a normal distribution t-N(μ t , σ t 2 ) with an expectation of μ t , σ t 2 ). Take μ t = 17.6, σ t 2 = 3.4. Then the probability density function of the electric vehicle return time t 0 is:

[0068]

[0069] The daily driving mileage S of electric vehicle users obeys a lognormal distribution S-log(μ s , σ s 2 ) with an expectation of μ s , σ s 2 ). Take μ s = 3.2, σ s 2 = 0.88. Then the probability density function of the daily driving mileage S of electric vehicles is:

[0070]

[0071] The charging duration required for an electric vehicle with a daily driving mileage of S is:

[0072]

[0073] Where: E is the power consumption per kilometer of the electric vehicle, with the unit of kWh / km. Take E = 0.5 kWh / km; P c is the charging power of the electric vehicle, with the unit of kW. Take P c = 3 kW; η c is the charging efficiency of the electric vehicle. Take ηc = 0.9;

[0074] According to the Monte Carlo method, the return time and daily driving mileage of a single electric vehicle are simulated, and the daily charging load of the electric vehicle is:

[0075]

[0076] Where: P ci is the charging power of the i-th electric vehicle, with the unit of kW, and I i is used to determine whether the i-th vehicle is charging at time t. I i = 0 means not charging, and I i = 1 means charging;

[0077] The state of charge (SOC) constraint of the electric vehicle is:

[0078] SOC min ≤ SOC ≤ SOC max

[0079] Where: SOC min , SOC max are the minimum and maximum values of the remaining power of the electric vehicle, respectively, with the unit of Ah;

[0080] Assume that the number of electric vehicles is 1000, and their charging behaviors are simulated; assume that after the daily travel of the electric vehicle ends, it starts to charge at a constant power until the state of charge (SOC) of the electric vehicle reaches 100%. During the charging period, it is not affected by the time-of-use electricity price incentive; the specific process of the unordered charging load prediction is as shown in the appendix of the specification Figure 1 As shown, after the prediction, the load curve of 1000 electric vehicles under unordered charging is obtained; the unordered charging load curve of the electric vehicle, the daily total load curve of the power grid, and the original load curve of the power grid are as shown in the appendix of the specification Figure 2 As shown, it can be seen that in the case of unordered charging of electric vehicles, the phenomenon of "peak on peak" appears in the total load of the power grid;

[0081] S2. Introduce the carbon trading mechanism into the unordered charging load prediction model of the electric vehicle in step S1, and calculate the carbon emissions and carbon quotas of power generation enterprises; assume that the load is provided by generator sets, and the baseline method is used to determine the free carbon emission quotas of the units; the carbon quota E obtained by the generator set at time t ct is as follows:

[0082] E ct = δ P P Mt

[0083] E Pt = c 1 P Mt

[0084] Among them: The active power output of the generator set satisfies P Mt = P t0 + P tEV for the power balance constraint; E ct is the carbon quota obtained at time t; E Pt is the carbon emission at time t; P Mt is the active power output of the generator set at time t, with the unit of kWh, and δ P is the carbon emission quota per unit of power generation, taking 0.798 kg / (kWh); c 1 is the power carbon emission factor, representing the CO 2 emission per degree of electricity, taking 0.8386 kg / (kWh);

[0085] S3. Introduce the reward coefficient ξ 1 and the carbon trading cost growth rate τ 1 into the conventional stepped carbon price to establish a reward and punishment type stepped carbon price model;

[0086] Introduce the reward coefficient ξ 1 into the conventional stepped carbon price to establish a reward and punishment type stepped carbon price model; when the total carbon emission of the power generation enterprise is greater than the free - allocated carbon quota, the power generation enterprise needs to purchase additional carbon emission rights through the carbon trading market for the generator set to continue generating power; as the load increases, the total carbon emission increases, and correspondingly, the carbon trading cost will also increase step - by - step under the reward and punishment type stepped carbon price model; when the total carbon emission of the power generation enterprise is less than the free - allocated carbon quota, the power generation enterprise can sell these surplus carbon emission rights to other enterprises in need of carbon emission rights through the carbon trading market to obtain additional income; as the load decreases, the total carbon emission also decreases, and correspondingly, the carbon trading cost will also decrease step - by - step under the reward and punishment type stepped carbon price model; when establishing the model, this method does not completely unify the sizes and numbers of the positive and negative carbon trading volume intervals because the carbon quota value at time t is not the median of the total carbon emission. Separately specifying the sizes and numbers of the positive and negative intervals can more effectively reflect the change of the carbon price; the formula for the reward and punishment type stepped carbon price is as follows:

[0087]

[0088] E t = E Pt - E ct

[0089] Among them: C t CO2 is the carbon trading cost at time t, with the unit of yuan, and λ 1 is the carbon trading cost base price, with the unit of yuan / ton; E tis the difference between the total carbon emissions and the free - allocated carbon quota at time t, with the unit of ton; τ 1 is the growth rate of the carbon trading cost for each stage; ξ 1 is the reward coefficient, L 0 is the length of the negative interval of the carbon cost, with the unit of ton, taking 40 tons; L 1 is the length of the positive interval of the carbon cost, with the unit of ton, taking 30 tons;

[0090] In the above reward - punishment carbon trading cost model, the relationship between the carbon trading price and the carbon trading volume can be seen more intuitively from the attached instructions of the specification Figure 3 When the carbon emissions are high, the carbon price rises step - by - step; when the carbon emissions are low, the salable carbon price also gradually increases;

[0091] S4. Establish a dynamic time - of - use electricity price model based on the reward - punishment stepped carbon price; the load of the distribution network fluctuates with time. Adjusting the electricity price according to the load change can more effectively and flexibly dispatch the charging behavior of electric vehicle users; at the same time, introduce the total electricity price growth coefficient a 1 and the total electricity price decline coefficient b 1 , and link the change of the initial electricity price with the base price of the carbon trading cost, so as to associate the electricity price change, the load change and the carbon trading cost change; respectively stipulate several electricity price rising intervals and electricity price falling intervals. The formula for the time - of - use electricity price at time t is as follows:

[0092]

[0093] X t0 =X 0 λ 1 / λ 0

[0094] a 1 =a 0 τ 1 / τ 0

[0095] b 1 =b 0 ξ 1 / ξ 0

[0096]

[0097] P max =max(P t0 +P tEV )

[0098] P min =min(P t0 +P tEV )

[0099] Among them: X 0 is the initial electricity price, taking 0.944 yuan; Y 0 is the electricity price decline interval, in MW, obtained by dividing the difference between the average daily total load and the minimum value by the number of increase intervals; Y 1 is the electricity price increase interval, obtained by dividing the difference between the maximum daily total load and the average value by the number of increase intervals, in MW; P av represents the average value of the daily total load; P t0 is the original grid load without electric vehicle load at time t, in MW; P tEV is the electric vehicle load at time t, in MW; T is the number of segments per day, taking 24; X t0 represents the initial electricity price under the influence of the reward and punishment ladder carbon price at time t; P max represents the maximum total load within a day, in MW; P min represents the minimum total load within a day, in MW; λ 0 is the initial carbon trading cost base price, taking 295 yuan / ton; τ 0 is the initial carbon trading cost growth rate, taking 0.3; a 1 represents the electricity price growth coefficient under the influence of the reward and punishment ladder carbon price; a 0 is the initial electricity price growth coefficient, with a value range of 0.05 - 0.389; ξ 1 is the initial reward coefficient, taking 0.2; b 1 represents the electricity price decline coefficient under the influence of the reward and punishment ladder carbon price; b 0 is the initial electricity price decline coefficient, with a value range of 0 - 0.1735;

[0100] S5. Establish an electric vehicle electricity price regulation model; combined with the attached Figures 4 - 5 to the specification, the consumption behavior of electric vehicle users will follow the change of electricity price, and users choose to accept the current price and change the charging time; therefore, the electric vehicle electricity price response curve is set as follows

[0101]

[0102] The electricity price response curve of electric vehicles is as shown in the attached Figure 4 to the specification. It can be clearly seen that when the electricity price is high, the number of electric vehicle chargings is relatively small, and when the electricity price is low, the number of electric vehicle chargings will increase significantly;

[0103] After the electricity price changes, the electric vehicle load power at time t is:

[0104] P tEVnew = ρP tEV

[0105] The total grid load after regulation is:

[0106] P 总 = P t0 + P tEVnew

[0107] where P tEVnew is the load power of the electric vehicle at time t after the electricity price changes; P 总 is the total grid load;

[0108] Considering from the aspects of the power grid company's profit and users, the dynamic time-of-use electricity price should not be lower than the marginal cost electricity price and a ceiling electricity price should also be set;

[0109] X min ≤ X t ≤ X max

[0110] where: X min , X max are the marginal cost electricity price and the maximum electricity price respectively, with the unit of yuan, taking 0.25 yuan and 2.5 yuan respectively;

[0111] S6. Establish a multi-objective optimization function for the orderly charging of electric vehicles; aiming at suppressing the total grid load, divide a day into T time periods, and take the minimum peak-valley mean square deviation of the superposition of the original grid load and the electric vehicle charging load as the optimization objective F 1 , in order to balance the interests of users, the power grid and power generation enterprises and achieve a win-win goal, the multi-optimization objective function considering the total charging cost and carbon trading cost respectively is:

[0112]

[0113] where: is the average value of the total grid load after dispatching through the time-of-use electricity price, with the unit of kWh; T is the number of segments of the daily time, taking 24 time periods in a day; F 1 is the objective function with the minimum peak-valley mean square deviation of the superposition of the original grid load and the electric vehicle charging load; F 2 is the objective function with the minimum total cost of electric vehicle charging users, with the unit of 10,000 yuan; F 3 is the objective function with the minimum carbon trading cost of power generation enterprises, with the unit of 10,000 yuan;

[0114] S7. Combine the attached instructions Figure 6 , and adopt the multi-objective non-dominated genetic NSGA-II solution algorithm; first, through the initial carbon trading cost base price λ 0 value, electricity price constraints and step 4, set the initial electricity price increase coefficient a 0 and the initial electricity price decrease coefficient b 0The upper and lower limits are used to achieve the purpose of electricity price constraint, and at the same time, the population size N and the original grid load P without the electric vehicle charging load are initialized. t0 The objective function F is obtained through steps S1 to S6. 1 ~F 3 The above steps are looped N times to obtain the objective functions of N populations. The empirical value of N is 100 - 300. The non-dominated sorting of the objective functions of N populations and the calculation of the crowding degree of population individuals are carried out. Then, it enters the population evolution cycle stage, and the objective function of the offspring is evaluated to obtain the three objective functions of the offspring. The parent population (N) and the offspring population (N) are merged (2N), and the non-dominated sorting and crowding degree calculation of the three objective functions of the merged population are carried out again. The new parent population (N) is obtained through the elite retention strategy until the evolution generation reaches the set maximum generation. After the iteration ends, the Pareto solutions obtained in each run are merged, and non-dominated sorting is applied to the merged Pareto solutions to screen out the Pareto optimal solutions with rank 1.

[0115] Embodiment 1:

[0116] Combined with the attached drawings of the specification Figures 7 - 8 , the unordered charging load of electric vehicles is generated by the Monte Carlo method, and the daily load data of a 10kV feeder area in a certain urban area is used as the basic data for example verification.

[0117] To verify the effectiveness of the proposed time-of-use electricity price electric vehicle orderly charging regulation strategy based on the carbon trading mechanism, let the carbon trading cost base price λ 1 be taken as 295 yuan / ton, the positive interval length L of the carbon cost 1 be taken as 30 tons, the negative interval length L of the carbon cost 0 be taken as 40 tons, the initial electricity price growth coefficient a 0 be taken as 0.389, the initial electricity price growth coefficient b 0 be taken as 0.1735, the carbon trading cost growth rate t 1 be taken as 0.3, the reward coefficient ξ 1 be taken as 0.20, and step S7 is looped 200 times.

[0118] And 3 scenarios are set for comparative analysis:

[0119] Scenario 1 is the scenario of electric vehicle charging optimal scheduling considering the reward and punishment type ladder carbon price.

[0120] Scenario 2 is the electric vehicle charging optimal scheduling considering only the non-reward ladder carbon price.

[0121] Scenario 3 is the situation of electric vehicle charging optimal scheduling considering the average carbon price and without reward.

[0122] Among them, the mean square error of Scenario 1 decreased by 53.46% and 60.68% compared with Scenarios 2 and 3 respectively; the user consumption of Scenario 2 decreased by 32.49% and 36.91% compared with Scenarios 1 and 3 respectively; the user consumption of Scenario 1 decreased by 6.55% compared with Scenario 3, and the carbon trading cost decreased by 10.99% and 25.37% compared with Scenarios 2 and 3 respectively; generally speaking, the time-of-use electricity price considering the reward and punishment-based stepped carbon price can conduct the orderly charging scheduling of electric vehicles more effectively; the scheduling results under the three scenarios are compared as follows in the table:

[0123] Scenario Mean square error / MW User consumption / 10,000 yuan Carbon cost / 10,000 yuan Scenario 1 977218.818 36.8324 10.6436 Scenario 2 2099658.956 24.8648 11.9583 Scenario 3 2485316.831 39.4131 14.2627

[0124] Combined with the impact of the reward and punishment-based stepped carbon price and the dynamic time-of-use electricity price on the charging behavior of electric vehicle users, analyze Table 1:

[0125] The user consumption in Scenario 1 is higher than that in Scenario 2

[0126] When considering the reward and punishment-based stepped carbon price, the time-of-use electricity price changes under the influence of the stepped carbon price and begins to schedule the charging behavior of electric vehicle users, causing most users to charge during the valley period and resulting in an increase in the valley area of the load curve. Compared with Scenario 2, the valley area of the load curve increases significantly. When Scenario 2 is in the valley area, it cannot profit from selling excess carbon emission rights, so the electricity price does not change because it is not affected by the change in the carbon price, and thus the valley of the load curve does not change. Therefore, compared with Scenario 1, only the curve in the peak area decreases, and the curve in the valley area does not rise, and users do not charge in large numbers, resulting in higher user consumption in Scenario 1 compared with Scenario 2;

[0127] The mean square error of Scenario 1 is higher than the other two scenarios;

[0128] In Scenario 1, the change range of the carbon price is large, and the stimulation to the electricity price is also correspondingly large. Therefore, when considering the reward and punishment-based stepped carbon price, the scheduling of electric vehicle users is better than the other three scenarios, thus significantly reducing the mean square error;

[0129] The carbon cost differences among the three scenarios

[0130] The reason why the carbon cost of Scenario 1 is lower than that of Scenario 2 is that under the reward mechanism, if the carbon emissions are lower than the carbon quota, the excess carbon quota can be sold for profit. However, in Scenarios 2 and 3, the excess carbon quota cannot be sold, making their carbon costs higher than those with the reward;

[0131] In the case of equal carbon price, since the price per ton of carbon emission rights does not increase with the increase in carbon emissions, the impact of carbon price on electricity price is weakened, which in turn leads to a weakened scheduling of the load curve, resulting in a higher load curve in the peak region compared to Scenario 2. Both Scenario 2 and Scenario 3 incorporate a carbon trading mechanism, but the difference between them lies in that when the carbon quota is greater than the carbon emissions, the carbon price in Scenario 2 rises step by step, while the carbon price in Scenario 3 remains unchanged. Therefore, in the peak region of Scenario 2, the scheduling is more significant, making the peak value lower, and ultimately resulting in a higher carbon cost in Scenario 3 compared to Scenario 2.

[0132] In summary, the optimized scheduling method for guiding the orderly charging of electric vehicles according to the present invention can effectively guide electric vehicle users to perform orderly charging behaviors by using the regulation method based on the reward and punishment type stepped carbon price and the dynamic time-of-use electricity price, thereby effectively achieving the effect of peak shaving and valley filling of the power grid and reducing the fluctuations of the power grid; using the change of the reward and punishment type stepped carbon price and the load to regulate the electricity price, and finally realizing the regulation of the charging behaviors of electric vehicle users.

[0133] The parts not detailed in the present invention are prior art. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting, aiming to include all changes falling within the meaning and scope of the equivalent elements within the present invention.

Claims

1. An optimized scheduling method for guiding the orderly charging of electric vehicles, characterized in that: It includes the following steps: S1. Establish an unordered charging load prediction model for electric vehicles based on the Monte Carlo method; S2. Introduce a carbon trading mechanism into the unordered charging load prediction model of electric vehicles in step S1, and calculate the carbon emissions and carbon quotas of power generation enterprises; E ct = δ P P Mt E Pt = c 1 P Mt Among them: E ct is the carbon quota obtained at time t; E Pt is the carbon emission at time t; P Mt is the active power output of the generator set at time t, with the unit of kW, and δ P is the carbon emission quota per unit of electricity generation, taking 0.798 kg / kWh; c 1 is the electricity carbon emission factor, representing the CO 2 emission generated per degree of electricity, taking 0.8386 kg / kWh; S3. Introduce the reward coefficient ξ 1 and the growth rate τ of carbon trading cost 1 to establish a reward and punishment-based stepped carbon price model; S4. Establish a dynamic time-of-use electricity price model based on a reward and punishment-based stepped carbon price; X t0 = X 0 λ 1 / λ 0 a 1 = a 0 τ 1 / τ 0 b 1 = b 0 ξ 1 / ξ 0 P max = max(P t0 + P tEV ) P min = min(P t0 + P tEV ) Where: X 0 is the initial electricity price; Y 0 is the electricity price decline range, in MW, obtained by dividing the difference between the average daily total load and the minimum value by the number of increase ranges; Y 1 is the electricity price increase range, obtained by dividing the difference between the maximum daily total load and the average value by the number of increase ranges, in MW; P av represents the average value of the daily total load; P t0 is the original grid load without electric vehicle load at time t, in MW; P tEV is the electric vehicle load at time t, in MW; T is the number of segments of each day's time; X t0 represents the initial electricity price under the influence of the reward and punishment type stepped carbon price at time t; P max represents the maximum total load within a day, in MW; P min represents the minimum total load within a day, in MW; λ 0 is the initial carbon trading cost base price; τ 0 is the initial carbon trading cost growth rate, taking 0.3; a 1 represents the electricity price growth coefficient under the influence of the reward and punishment type stepped carbon price; a 0 is the initial electricity price growth coefficient, with a value range of 0.05 to 0.389; ξ 1 is the initial reward coefficient, taking 0.2; b 1 represents the electricity price decline coefficient under the influence of the reward and punishment type stepped carbon price; b 0 is the initial electricity price decline coefficient, with a value range of 0 to 0.1735; S5. Establish an electricity price regulation model for electric vehicles; P tEVnew = ρP tEV P 总 = P t0 + P tEVnew Among them, P tEVnew is the load power of the electric vehicle at time t after the electricity price changes; P 总 is the total grid load; S6. Establish a multi-objective optimization function for the orderly charging of electric vehicles; Wherein: is the average value of the total grid load after dispatching by time-of-use electricity price, with the unit of kW; T is the number of segments of each day's time; F 1 is the objective function with the minimum peak-valley mean square deviation of the superposition of the original grid load and the electric vehicle charging load; F 2 is the objective function with the minimum total cost of electric vehicle charging users, with the unit of ten thousand yuan; F 3 is the objective function with the minimum carbon trading cost of power generation enterprises, with the unit of ten thousand yuan; S7. Adopt the multi-objective non-dominated genetic NSGA-II solution algorithm; obtain the objective function F through steps S1 to S6 1 ~F 3 ; loop the above steps N times to obtain the objective functions of N populations; perform non-dominated sorting on the objective functions of N populations and calculate the crowding degree of population individuals; then enter the population evolution loop stage, evaluate the objective functions of the offspring, and obtain the three objective functions of the offspring; merge the parent population (N) and the offspring population (N) (2N), perform non-dominated sorting and crowding degree calculation on the three objective functions of the merged population again, and obtain the new parent (N) through the elitist retention strategy until the number of generations of evolution reaches the set maximum number of generations; after the iteration ends, merge the Pareto solutions obtained in each run, apply non-dominated sorting to the merged Pareto solutions, and screen out the Pareto optimal solutions with rank 1; In step S1, the return time t of daily residential users is set 0 subject to a normal distribution with an expectation of μ t and a variance of σ t 2 t - N(μ t , σ t 2 ), taking μ t = 17.6 and σ t 2 = 3.4, the probability density function of the electric vehicle return time t 0 is as follows: The daily driving mileage S of electric vehicles follows a lognormal distribution with an expected value of μ s , variance of σ s 2 S - log(μ s , σ s 2 ); taking μ s = 3.2, σ s 2 = 0.88, the probability density function of the daily driving mileage S of electric vehicles is: The charging duration required for an electric vehicle with a daily driving mileage of S is: Where: E is the power consumption per kilometer of the electric vehicle, with the unit of kWh / km; P c is the charging power of the electric vehicle, with the unit of kW; η c is the charging efficiency of the electric vehicle, and take η c = 0.9; Based on the Monte Carlo method, the return time and daily driving mileage of a single electric vehicle are simulated, and the daily charging load of the electric vehicle is: Where: P ci is the charging power of the i-th electric vehicle, with the unit of kW, I i is used to determine whether the i-th vehicle is charging at time t, I i = 0 indicates not charging, I i = 1 indicates charging; The reward and punishment-based stepped carbon price model in step S3 is: E t = E Pt - E ct Among them: C t CO2 is the carbon trading cost at time t, with the unit of yuan, and λ 1 is the base price of carbon trading cost, with the unit of yuan / ton; E t is the difference between the total carbon emissions and the free-allocated carbon quota at time t, with the unit of ton; τ 1 is the growth rate of carbon trading cost for each stage; ξ 1 is the reward coefficient, L 0 is the length of the negative interval of carbon cost, with the unit of ton; L 1 is the length of the positive interval of carbon cost, with the unit of ton.

2. The optimized scheduling method for guiding the orderly charging of electric vehicles according to claim 1, characterized in that: The state of charge SOC constraint of the electric vehicle in step S1 is: SOC min ≤SOC≤SOC max where: SOC min , SOC max are the minimum and maximum values of the remaining power of the electric vehicle, respectively, with the unit of Ah.

3. The optimized scheduling method for guiding the orderly charging of electric vehicles according to claim 1, characterized in that: In step S2, the active power output of the generator set satisfies P Mt = P t0 + P tEV for the power balance constraint.

4. The optimized scheduling method for guiding the orderly charging of electric vehicles according to claim 1, characterized in that: The electricity price constraint in step S5 is: X min ≤X t ≤X max Where: X min , X max are the marginal cost electricity price and the highest electricity price respectively, with the unit of yuan.

5. The optimized scheduling method for guiding the orderly charging of electric vehicles according to claim 1, characterized in that: In step S7, first, through the initial carbon trading cost base price λ 0 whose value is the initial electricity price, set the upper and lower limits of the initial electricity price growth coefficient a 0 and the initial electricity price decline coefficient b 0 to achieve the purpose of electricity price constraint. At the same time, initialize the population size N and the original power grid load P without the electric vehicle charging load t0 .

Images