A multi-objective online scheduling method for an electric vehicle charging station

Abstract

The present invention discloses a multi-objective online scheduling method for an electric vehicle charging station, belonging to the technical field of charging behavior scheduling of electric vehicle charging stations. The present invention proposes an online scheduling framework for an electric vehicle charging station considering multiple objectives, dynamically adjusts the weight values corresponding to each objective based on the pre-established objectives and historical data, and realizes the online solution of the compromise solution to the multi-objective electric vehicle charging station scheduling problem. And in order to achieve online optimization, the present invention proposes a definition method for the service quality of an electric vehicle charging station based on the arrival curve and the departure curve, and realizes the online calculation of the service quality. In the framework, the feasible region is adjusted online, realizing the real-time response to the uncertain environment, reducing the operating cost of the system while improving the service quality. The present invention can flexibly and quickly give a scheduling scheme according to environmental changes, reducing the power supply cost of power suppliers while ensuring the service quality of electric vehicle charging stations.

Description

Technical Field

[0001] The present invention belongs to the technical field of charging behavior scheduling for electric vehicle charging stations, and more specifically, relates to a multi-objective online scheduling method for electric vehicle charging stations. Background Art

[0002] To protect the environment, electric vehicles (EVs) have been widely used in our society. It is predicted that by 2030, the number of electric vehicles worldwide will exceed 100 million. However, with the increase in the number of electric vehicles, due to the lack of sufficient public charging piles, it is difficult to meet the charging demand in a timely manner. Worldwide in 2021, each public charging pile had to serve up to 20 electric vehicles. Therefore, using limited charging piles to meet the charging demand of electric vehicles has attracted people's attention.

[0003] To address this issue, optimizing the charging operation of electric vehicle charging stations is an effective method. However, the charging demand of electric vehicle charging stations may be a difficult quantity to predict. For this reason, using an online optimization method to optimize the charging operation of electric vehicle charging stations is an effective scheduling means. However, existing work only performs single-objective scheduling using the online optimization method. The charging operation of electric vehicles has the nature of multiple objectives, mainly due to the various needs of multiple stakeholders. Therefore, in some studies, the charging operation of electric vehicle charging stations (EVCS) is considered a multi-objective optimization problem.

[0004] When using an offline multi-objective optimization method to schedule electric vehicle charging stations, it is generally scheduled based on the predicted demand. Generally speaking, the prediction is based on historical data. However, if the amount of historical data is insufficient to support the prediction, the traditional scheduling based on predicted data cannot be carried out; if the prediction accuracy is insufficient or the predicted data has a large gap with the real data (for example, the predicted value is significantly too large or too small within a certain time period), it will also lead to serious distortion of the scheduling result. Therefore, aiming at this problem, how to achieve multi-objective scheduling of electric vehicle charging stations without relying on predicted data is a difficult problem. Summary of the Invention

[0005] In view of the above defects or improvement requirements of the prior art, the present invention provides a multi-objective online scheduling method for electric vehicle charging stations, aiming to perform multi-objective optimal scheduling of electric vehicle charging stations without relying on predicted demand data, while minimizing the charging cost while quickly meeting the charging demand of electric vehicles.

[0006] To achieve the above object, the present invention provides a multi-objective online scheduling method for an electric vehicle charging station, including:

[0007] S1. Construct a multi-objective online scheduling model; the multi-objective online scheduling model takes the weighted sum of the service quality of all electric vehicle charging stations and the total cost of the power supplier within a time interval t as the objective function; wherein, the service quality of the electric vehicle charging station is the difference between the arrival curve and the service curve; the weight of the power supplier is the average difference between the total cost within the time interval t and the corresponding set target; the weight of the electric vehicle charging station i is the average difference between the service quality within the time interval t and the corresponding set target.

[0008] S2. The electric vehicle charging station and the power supplier respectively set their scheduling targets and initialize their respective weights.

[0009] S3. When an electric vehicle is connected to a charging pile, calculate the charging pile arrival curve according to its charging demand; the electric vehicle charging station i calculates its current arrival curve value and calculates the power demand range.

[0010] S4. Take the current weight values and power demand ranges of each electric vehicle charging station as inputs, solve the multi-objective online scheduling model within the time interval t to obtain the power supply power provided by the power supplier; the power supplier supplies power to the electric vehicle charging station, and the electric vehicle charging station distributes the power to each electric vehicle charging pile.

[0011] S5. The electric vehicle charging station calculates its current arrival curve and service quality; the power supplier calculates its total cost; the electric vehicle charging station and the power supplier respectively calculate their weights at the next moment. If the scheduling period T is not reached, return to execute step S3; if the scheduling period T is reached, end the scheduling.

[0012] Further, the objective function of the multi-objective online scheduling model is:

[0013]

[0014] α T (t) is the weight value of the power supplier, is the weight value of the electric vehicle charging station i, , debt T (t) and debt i (t) are respectively the gap values from the target of the power supplier and the electric vehicle charging station i within the time interval t;

[0015] QoS i (t) is the service quality value of the electric vehicle charging station i within the time interval t, QoS i (t) = (A i (t) - D i(t))Δt, A i D(t) is the arrival curve value of electric vehicle charging station i during the time interval t i D(t) is the departure curve value of electric vehicle charging station i during the time interval t; is the total cost of the power supplier during the time interval t; (·) + represents max(0, ·); is the number of electric vehicle charging stations; Δt is the length of the time interval.

[0016] Furthermore, the arrival curve value of electric vehicle charging station i during the time interval t is:

[0017]

[0018] A j A(t) is the arrival curve value of electric vehicle charging pile j during the time interval t:

[0019]

[0020] E re,j E(t) is the cumulative energy demand of charging pile j from the start of scheduling to the time interval t; is the maximum charging power of charging pile j.

[0021] Furthermore, the departure curve value of electric vehicle charging station i during the time interval t is:

[0022]

[0023] D j D(t) is the departure curve value of electric vehicle charging pile j during the time interval t:

[0024]

[0025] P ev,j P(t) is the actual charging power of charging pile j during the time interval t.

[0026] Furthermore, the power demand range of electric vehicle charging station i during the time interval t is:

[0027]

[0028]

[0029] and are the upper and lower limits of the energy demand of electric vehicle charging station i during the time interval t, P ev,j P(t) is the actual charging power of charging pile j during the time interval t, is the lower limit of the charging power of charging pile j.

[0030] Furthermore, the constraint conditions of the multi-objective online scheduling model include:

[0031] Charging constraint of the electric vehicle charging station:

[0032]

[0033] Power supply constraint of the power supplier:

[0034]

[0035] Supply-demand balance constraint:

[0036]

[0037] Among them and are respectively the upper and lower limits of the power supply of the power supplier, P T (t) is the power supply power of the power supplier in the time interval t, P i (t) is the total charging power of the electric vehicle charging station i in the time interval t, and φ(·) is the power supply loss function of the power supplier.

[0038] Furthermore, the total cost of the power supplier in the time interval t

[0039] C T (t) is the power generation cost of the power supplier in the time interval t, C T (t) = aP T (t) 2 +bP T (t)+c, where a, b, and c are all cost coefficients of the generator set, P T (t) is the power supply power of the power supplier in the time interval t, F(P i (t)) is the power transmission cost to the electric vehicle charging station i in the time interval t, F(P i (t)) = ∈d i P i (t), ∈ is the grid usage cost per unit distance, d i is the electrical distance when transmitting power to the electric vehicle charging station i, and P i (t) is the total charging power of the electric vehicle charging station i in the time interval t.

[0040] Generally speaking, compared with the prior art, the above technical solutions conceived by the present invention can achieve the following beneficial effects.

[0041] Compared with the scheduling of traditional offline multi-objective electric vehicle charging stations, the present invention combines online optimization with multi-objective optimization. According to the Target based online dynamic weighted algorithm (TBODWA), an online scheduling framework for electric vehicle charging stations considering multiple objectives is proposed. Based on the pre-established objectives and historical data, the weight values corresponding to each objective are dynamically adjusted, realizing the online solution of the compromise solution to the multi-objective electric vehicle charging station scheduling problem. And in order to achieve online optimization, the present invention proposes a definition and measurement method for the service quality of electric vehicle charging stations based on arrival curves and departure curves, realizing the online calculation of service quality. The feasible region in the framework is adjusted online, thus realizing the real-time response to the uncertain environment, while improving the service quality and reducing the operating cost of the system. The designed online scheduling framework for electric vehicle charging stations considering multiple objectives can flexibly and quickly give a scheduling scheme according to environmental changes, and can reduce the power supply cost of power suppliers while ensuring the service quality of electric vehicle charging stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic structural diagram of the online scheduling framework for electric vehicle charging stations considering multiple objectives provided by the present invention.

[0043] Figure 2 Among them, (a)-(d) are schematic diagrams of the calculation method for the service quality of the electric vehicle charging station provided by the present invention.

[0044] Figure 3 It is the process intention of the online multi-objective optimization algorithm TBOWA provided by the present invention.

[0045] Figure 4 It is the convergence curve provided by the embodiment of the present invention under Target 1 (O T =0, O1 = 0, O2 = 0, O3 = 0), where (a) is the average service quality curve of the electric vehicle charging station, and (b) is the average cost curve of the power supplier.

[0046] Figure 5 It is the convergence curve provided by the embodiment of the present invention under Target 2 (O T =80, O1 = 40, O2 = 100, O3 = 15), where (a) is the average service quality curve of the electric vehicle charging station, and (b) is the average cost curve of the power supplier.

[0047] Figure 6 It is the convergence curve provided by the embodiment of the present invention under Target 3 (O TConvergence curves under = 65, O1 = 17, O2 = 37, O3 = 3), where (a) is the average service quality curve of the electric vehicle charging station and (b) is the average cost curve of the power supplier.

[0048] Figure 7 For the weight change curves provided by the embodiments of the present invention under Target 4-6 (Target 4: O T = 35, O1 = 18, O2 = 36, O3 = 3, Target 5: O T = 35, O1 = 13.9621, O2 = 12.9839, O3 = -9.0960, Target 6: O T = 35, O1 = 9.9242, O2 = -10.0322, O3 = -21.1920), where (a) is the weight curve of the service quality of the electric vehicle charging station and (b) is the weight curve of the cost of the power supplier. Detailed implementation manners

[0049] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0050] Referring to Figure 1 , a multi-objective online scheduling method for an electric vehicle charging station provided by the present invention includes:

[0051] S1. Construct a multi-objective online scheduling model; the multi-objective online scheduling model takes the minimum weighted sum of the service quality of all electric vehicle charging stations and the total cost of the power supplier within a time interval t as the objective function; wherein, the service quality of the electric vehicle charging station is the difference between the arrival curve and the service curve; the weight of the power supplier is the average difference between the total cost within the time interval t and the corresponding set target; the weight of the electric vehicle charging station i is the average difference between the service quality within the time interval t and the corresponding set target;

[0052] S2. The electric vehicle charging station and the power supplier respectively set their scheduling targets O T and O i , and initialize their respective weights; wherein, the target of the electric vehicle charging station is the average service quality, and the target of the power supplier is the average cost of power generation and supply;

[0053] S3. When an electric vehicle is connected to a charging pile, calculate the arrival curve of the charging pile according to its charging demand; the electric vehicle charging station i calculates its current arrival curve value and calculates the power demand range;

[0054] S4. Take the current weight values and power demand ranges of each electric vehicle charging station as inputs, and solve the multi-objective online scheduling model within the time interval t to obtain the power supply provided by the power supplier; the power supplier supplies power to the electric vehicle charging station, and the electric vehicle charging station distributes power to each electric vehicle charging pile;

[0055] S5. The electric vehicle charging station calculates its current arrival curve and quality of service; the power supplier calculates its total cost; the electric vehicle charging station and the power supplier calculate their weights at the next moment respectively. If the scheduling period T is not reached, return to step S3; if the scheduling period T is reached, end the scheduling.

[0056] The overall optimized scheduling model is:

[0057]

[0058]

[0059] D j (t) ≤ E re,j (t)

[0060]

[0061]

[0062] where is the quality of service of all electric vehicle charging stations within the time interval t, that is:

[0063]

[0064] is the number of electric vehicle charging stations; is the total cost of the power supplier within the time interval t; T is the scheduling period.

[0065] The present invention uses the arrival curve and departure curve to conveniently characterize the charging behavior of electric vehicle charging stations online, providing a basis for online scheduling modeling of electric vehicle charging stations.

[0066] Such as Figure 2 in (a), the arrival curve of electric vehicle charging pile j is:

[0067]

[0068] E re,j (t) is the cumulative energy demand of charging pile j from the start of scheduling to the time interval t, is the maximum charging power of charging pile j, and Δt is the length of the time interval;

[0069] The arrival curve of the electric vehicle charging station i is obtained by summing the arrival curves of all its electric vehicle charging piles, i.e.:

[0070]

[0071] The arrival curve reflects the charging curve under the condition that the electric vehicle charging station can fastest meet the charging demand of electric vehicles. Opposite to the arrival curve is the departure curve. The departure curve of the electric vehicle charging pile is as shown in Figure 2 (b) in, i.e.:

[0072]

[0073] P ev,j (t) is the actual charging power of charging pile j in the time interval t;

[0074] Similar to the arrival curve, the departure curve of the electric vehicle charging station i is obtained by summing the departure curves of all its electric vehicle charging piles, i.e.:

[0075]

[0076] The quality of service of the electric vehicle charging station i is the area between the arrival curve and the departure curve, as shown in Figure 2 (c) in, i.e.:

[0077]

[0078] QoS i is the overall cycle quality of service of the electric vehicle charging station i. By discretizing the quality of service, as shown in Figure 2 (d) in, the calculation method of the quality of service value of the electric vehicle charging station i in the time interval t can be obtained as:

[0079] QoS i (t) = (A i (t) - D i (t))Δt

[0080] A i (t) is the arrival curve value of the electric vehicle charging station i in the time interval t, and D i (t) is the value of the departure curve of the electric vehicle charging station i in the time interval t.

[0081] The method for calculating the quality of service of the electric vehicle charging station proposed by the present invention can calculate the quality of service of the electric vehicle charging station at the current moment online, providing a basis for online optimization.

[0082] Among the constraints, from top to bottom are the charging power constraint, the charging amount constraint, the power generation constraint, and the supply-demand balance constraint. Among them, φ(P T(t) is the power loss of the power supplier, and the calculation method is:

[0083]

[0084] is the loss coefficient, P T (t) is the power supply power of the power supplier in the time interval t; obviously, in the case of adopting this loss function, the supply-demand balance constraint is a non-convex constraint. Therefore, the supply-demand balance constraint is modified to:

[0085]

[0086] P i (t) is the total charging power of the electric vehicle charging station i in the time interval t;

[0087] By adjusting the supply-demand balance constraint, the original non-convex problem is transformed into a convex problem, which facilitates the solution of the problem.

[0088] Combining the charging power constraint and the charging amount constraint, it can be known that for any electric vehicle charging pile j, the departure curve can never reach above the curve again, that is

[0089]

[0090] Obviously, this constraint also holds for any electric vehicle charging station i, that is:

[0091]

[0092] Combining the definitions of the arrival curve and the departure curve and transforming this constraint, the upper and lower limits of the charging power of the electric vehicle charging station i in the time interval t can be obtained as follows:

[0093]

[0094]

[0095] and are the upper and lower limits of the energy demand of the electric vehicle charging station i in the time interval t, is the lower limit of the charging power of the charging pile j;

[0096] Using the dynamic weighted sum method, in each time interval t, the multi-objective optimization problem is transformed into the following single-objective optimization problem: Assume that the weights corresponding to the electric vehicle charging station i and the power supplier in the time interval t are and α T (t). This weight is calculated in the previous time interval.

[0097] The objective function is:

[0098]

[0099] The constraints include:

[0100] Charging constraints for electric vehicle charging stations:

[0101]

[0102] Power supply constraints of power suppliers:

[0103]

[0104] Supply-demand balance constraints:

[0105]

[0106] The present invention solves this sub-problem at each time interval t, and obtains a compromise solution of the multi-objective optimization problem through the previously obtained weight values, thereby achieving the optimal operation in the current state and seeking its compromise solution on the premise of ensuring economic indicators and service quality.

[0107] The calculation method of the weight values of the power supplier and the electric vehicle charging station i is as Figure 3 shown:

[0108]

[0109]

[0110] Among them, the difference values between the power supplier and the electric vehicle charging station i and the objectives are respectively:

[0111]

[0112] debt i (t) = QoS i (t) - O i

[0113] By calculating the average difference value between the current objectives and the objectives, the difference between the current performance of the framework and the pre-set objective values is measured, and the weight values of different objectives in the next optimization are further determined, realizing the dynamic adjustment of the objective weights, so as to achieve the balance between various objective values considering the environmental uncertainty.

[0114] The total cost of the power supplier at time interval t C T (t) is the power generation cost of the power supplier at time interval t, C T (t) = aP T (t) 2+bP T (t)+c, where a, b, and c are all cost coefficients of the generator set, and F(P i (t)) is the transmission cost to the electric vehicle charging station i during the time interval t, and F(P i (t)) = ∈d i P i (t), ∈ is the grid usage cost per unit distance, and d i is the electrical distance when transmitting power to the electric vehicle charging station i.

[0115] Based on the proposed framework, an embodiment is set with six groups of target values as follows:

[0116] Target 1: O T = 0, O1 = 0, O2 = 0, O3 = 0;

[0117] Target 2: O T = 80, O1 = 40, O2 = 100, O3 = 15;

[0118] Target 3: O T = 65, O1 = 17, O2 = 37, O3 = 3;

[0119] Target 4: O T = 35, O1 = 18, O2 = 36, O3 = 3;

[0120] Target 5: O T = 35, O1 = 13.9621, O2 = 12.9839, O3 = -9.0960;

[0121] Target 6: O T = 35, O1 = 9.9242, O2 = -10.0322, O3 = -21.1920.

[0122] The electric vehicle demand data used in the embodiment comes from the ACN dataset. There are 3 electric vehicle charging stations, which respectively contain 54, 52, and 8 single-phase AC electric vehicle charging piles, with a maximum output power of 7 kW and a minimum output power of 0. The electrical distances between the power supplier and the electric vehicle charging stations are d1 = 1.00, d2 = 3.72, and d3 = 3.77. The transmission cost per unit electrical distance and unit power ∈ = 0.2. The power generation related parameters of the power supplier are shown in Table 1.

[0123] Table 1 Power generation related parameters of the power supplier

[0124]

[0125] The scheduling result in this embodiment is as Figure 4In (a) and (b)~ Figure 7 as shown in (a) and (b) of the figure. The scheduling results under each target are shown in Table 2.

[0126] Table 2 Scheduling Results

[0127]

[0128]

[0129] Target1 is an overly high indicator without considering the balance between different targets. Target 2 is an easily achievable target value. Target 3 is a target that is relatively close to the Pareto front and is relatively difficult to achieve. It can be seen from Table 2 that since Target 1 does not fully consider the balance between targets, although the cost of the power supplier is low, the service quality of each electric vehicle charging station is high. Since Target 2 is an easily achievable target value, the corresponding scheduling results have all exceeded the preset target value. The optimization result corresponding to Target 3 is relatively close to the preset target. The scheduling results under the above three target values can verify the effectiveness of the framework proposed by the present invention.

[0130] Target4 is the target to be achieved, and the other two Targets have the same target-based optimal solution as that corresponding to Target4. Target5 and Target6 are used to accelerate convergence. In other words, all the optimization results with Target4 - 6 are expected to converge to Target4 because they have the same target-based optimal solution. Setting Target4 - 6 is to verify the influence of different degrees of strictness of the target on convergence. It can be seen from the scheduling results in Table 2 that the stricter the Target, the better the convergence result.

[0131] The last column of Table 2 is the average time required to solve the sub-problem in each time interval. It can be seen that the average time required to solve the sub-problem in each time interval is within 0.1 s, meeting the requirement of timeliness in online scheduling.

[0132] Those skilled in the art can easily understand that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A multi-objective online scheduling method for an electric vehicle charging station, characterized in that, Including: S1. Construct a multi-objective online scheduling model; The multi-objective online scheduling model aims to minimize the weighted sum of the service quality of all electric vehicle charging stations and the total cost of the power supplier within a time interval t ; where the service quality of the electric vehicle charging station is the difference between the arrival curve and the departure curve; the weight of the power supplier is the average difference between the total cost within the time interval t and the corresponding set target; the weight of the electric vehicle charging station i is the average difference between the service quality within the time interval t and the corresponding set target. S2. The electric vehicle charging station and the power supplier respectively set their scheduling objectives and initialize their respective weights; S3. When the electric vehicle is connected to the charging pile, calculate the charging pile arrival curve according to its charging requirements; for the electric vehicle charging station i Calculate its current arrival curve value and calculate the power demand range; S4. Taking the current weight values and power demand ranges of each electric vehicle charging station as inputs, within the time interval t solve the multi-objective online scheduling model to obtain the power supply provided by the power supplier; the power supplier supplies power to the electric vehicle charging stations, and the electric vehicle charging stations distribute power to each electric vehicle charging pile; S5. The electric vehicle charging station calculates the current arrival curve and service quality; the power supplier calculates its total cost; the electric vehicle charging station and the power supplier respectively calculate the weights for the next moment. If the scheduling period T is not reached, return to execute step S3; if the scheduling period T is reached, end the scheduling; The objective function of the multi-objective online scheduling model is: is the weight value of the power supplier, , is the weight value of the electric vehicle charging station i . , and are the gap values between the power supplier and the electric vehicle charging station i and the target during the time interval t respectively; For an electric vehicle charging station i The quality of service value during the time interval t , , For an electric vehicle charging station i The arrival curve value during the time interval t , For an electric vehicle charging station i The departure curve value during the time interval t ; Is the total cost of the power supplier during the time interval t ; Represents ; Is the number of electric vehicle charging stations; Is the length of the time interval; Electric vehicle charging station i At the time interval t The arrival curve value is: For an electric vehicle charging pile j Reach the curve value within the time interval t as follows: For the charging pile j The cumulative energy demand from the start of scheduling to the time interval t ; For the charging pile j The maximum charging power; Electric vehicle charging station i In the time interval t The departure curve value is: For an electric vehicle charging pile j In the time interval t Deviating from the curve value: For a charging pile j During the time interval t Actual charging power 2. The multi-objective online scheduling method for an electric vehicle charging station according to claim 1, characterized in that Electric vehicle charging station i In the time interval t The power demand range is: and are respectively the upper and lower limits of the energy demand of the electric vehicle charging station i in the time interval t . is the actual charging power of the charging pile j in the time interval t . is the lower limit of the charging power of the charging pile j .

3. The multi-objective online scheduling method for an electric vehicle charging station according to claim 2, wherein, The constraint conditions of the multi-objective online scheduling model include: Charging constraints of the electric vehicle charging station: Power supply constraints of the power supplier: Supply-demand balance constraints: wherein and are respectively the upper and lower limits of the power supply quantity of the power supplier, is the power supply power of the power supplier in the time interval t ; is the total charging power of the electric vehicle charging station i in the time interval t ; is the power supply loss function of the power supplier.

4. The multi-objective online scheduling method of an electric vehicle charging station according to claim 3, wherein, Time interval t Total cost of the power supplier is the time interval t The power generation cost of the power supplier , are all cost coefficients of the generator sets is the power supply power of the power supplier during the time interval t is the time interval t to the electric vehicle charging station i transmission cost , is the grid usage cost per unit distance is for the electric vehicle charging station i electrical distance during transmission is the electric vehicle charging station i during the time interval t total charging power​

Images