Human-in-the-loop electricity carbon collaborative optimization method for electric vehicle charging resources

Abstract

The invention belongs to the technical field of power system optimization, and discloses a human-in-the-loop power carbon collaborative optimization method for electric vehicle charging resources, which comprises the following steps of: acquiring multi-source heterogeneous data including power supply side data, network side data and load side data; based on an economic dispatching principle, identifying a marginal unit in the power grid at the current moment, and calculating according to a carbon emission coefficient of the marginal unit to obtain a marginal carbon emission factor of the power grid; correcting the marginal carbon emission factor by using network side data to obtain an effective marginal carbon emission factor; calculating the carbon emission of the charging behavior of the electric vehicle based on the effective marginal carbon emission factor, calculating the operation cost according to the load side data, and predicting the adoption probability of a user to a charging scheme through a logistic regression model to calculate the user deviation; the problem that in the prior art, carbon emission accounting precision is insufficient, spatial difference is neglected, and a real-time feedback adjusting mechanism is lacked, so that collaborative optimization of degree resources and power grid low-carbon operation is difficult to achieve is effectively solved.

Description

Technical Field

[0001] This invention belongs to the field of power system optimization technology, specifically relating to a human-environment-electricity-carbon synergistic optimization method for electric vehicle charging resources. Background Technology

[0002] With the explosive growth in the number of electric vehicles, the impact of their charging load on urban power grids is becoming increasingly significant. Traditional disorderly charging not only exacerbates the peak-valley difference in the power grid, but may also lead to concentrated charging during periods of high carbon emission intensity in the power grid, which contradicts the low-carbon original intention of transportation electrification.

[0003] In the field of electric vehicle charging resource management, existing technologies are primarily based on the following well-known principles and methods: Power systems meet real-time load demands through economic dispatch following a "merit order" principle, i.e., dispatching generators in ascending order of marginal cost. Correspondingly, carbon emission accounting typically uses the grid's "average carbon emission factor" based on historical statistics. Optimization of electric vehicle charging load mainly relies on price signals guiding time transitions, and dispatch is achieved through centralized or distributed algorithms. The "Human-in-the-Loop" design concept, aimed at improving the adaptability and acceptability of complex systems, has gained importance in related system designs, aiming to incorporate human decision feedback into automatic control loops.

[0004] Currently, electric vehicle charging optimization methods suffer from the following problems: First, there is insufficient understanding of the electricity-carbon coupling mechanism. Existing systems mostly use static or average carbon emission factors, which cannot reflect real-time marginal carbon emission changes. Second, spatial differences in transmission and distribution networks are ignored. Traditional models assume the power grid is a single node and do not consider the impact of transmission losses on carbon emissions. Third, there is a lack of a "human-in-the-loop" feedback mechanism. Existing optimization systems are mostly open-loop decision-making systems, unable to dynamically adjust target weights according to the actual preferences of operators, resulting in system rigidity and low adoption rates. Therefore, existing technologies suffer from insufficient accuracy in carbon emission accounting, neglect of spatial differences, and a lack of real-time feedback and adjustment mechanisms, making it difficult to achieve coordinated optimization of power resources and low-carbon power grid operation. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the present invention aims to provide a human-centered, environmental, and carbon-coordinated optimization method for electric vehicle charging resources. This method solves the problems of insufficient carbon emission accounting accuracy, neglect of spatial differences, and lack of real-time feedback and adjustment mechanisms in existing technologies, which make it difficult to achieve coordinated optimization of energy resources and low-carbon operation of the power grid.

[0006] The objective of this invention can be achieved through the following technical solutions: A method for co-optimizing the energy, environment, and carbon emissions of electric vehicle charging resources includes the following steps: Acquire multi-source heterogeneous data, including power supply side data, network side data, and load side data; Based on the principle of economic dispatch, marginal units in the power grid at the current moment are identified, and the marginal carbon emission factor of the power grid is calculated based on the carbon emission coefficient of the marginal units. By using network-side data to correct the marginal carbon emission factor, an effective marginal carbon emission factor can be obtained. The carbon emissions of electric vehicle charging behavior are calculated based on the effective marginal carbon emission factor, and the operating costs are calculated based on load-side data. The probability of users adopting charging solutions is predicted by a logistic regression model to calculate user bias. The carbon emissions, operating costs, and user bias are normalized, and an objective function is constructed to minimize the weighted sum of carbon emissions, operating costs, and user bias. Constraints are established based on system operation, charging facility capacity, and user demand. Based on the objective function and constraints, a joint optimization solution is performed to obtain the optimal charging scheme for each charging station. After the optimized charging plan is executed in each scheduling cycle, the weighting coefficients of carbon emissions, operating costs and user deviation in the objective function for the next scheduling cycle are updated based on the management personnel's score of the execution results.

[0007] Furthermore, the power supply side data includes the real-time output power of each generator set, the upper and lower limits of the technical output of each generator set, the marginal operating cost quotation, and the carbon emission coefficient; Network-side data includes the power grid topology, voltage phasors at each node, and line resistance parameters; Load-side data includes the real-time charging power of each charging station, the state of charge of the batteries of each vehicle being charged, the estimated dwell time, and the user's historical charging behavior data.

[0008] Furthermore, the marginal carbon emission factor of the power grid is calculated based on the carbon emission coefficient of the marginal generating units, and the specific calculation formula is as follows: in, Indicates the marginal carbon emission factor; Represents a set of generator sets; i Indicates generator set; Indicates generator set i The carbon emission coefficient; This indicates that the total load of the power grid is Under the operating conditions, the generator set i At the present moment t The output power is And only marginal units at the current moment t Output power It is a non-zero value.

[0009] Furthermore, the specific calculation formula for correcting the marginal carbon emission factor using network-side data is as follows: in, Indicates the effective marginal carbon emission factor; Indicates the marginal transmission loss rate; V Represents the node voltage phasor; R This indicates the line resistance parameter.

[0010] Furthermore, the carbon emissions from electric vehicle charging are calculated based on the effective marginal carbon emission factor, using the following formula: in, This represents the total carbon emissions generated by all scheduled electric vehicle charging activities. , This represents a charging scheme vector, indicating each charging station. s During the period t Power and location selection; The specific formula for calculating operating costs is as follows: in, Indicates operating costs; Indicates the charging price; The mathematical expression for the logistic regression model is as follows: in, The probability of users adopting a charging solution; The price discount represents the percentage difference between the actual electricity price in the optimized charging plan and the preset benchmark electricity price for that period. The carbon emission reduction margin represents the proportion of carbon emission reduction achieved by users charging according to the optimized charging plan compared to unordered charging. The convenience score is calculated based on the time and space convenience of the optimized charging solution. User bias is calculated based on the probability of users adopting charging solutions. The specific calculation formula is as follows: in, Indicates user bias; This indicates that, given a charging scheme vector Under these conditions, users u Probability of adopting a charging solution.

[0011] Furthermore, an objective function is constructed to minimize the weighted sum of carbon emissions, operating costs, and user bias, with the specific expression as follows: in, Represent the objective function; , , These are the normalized operating costs, carbon emissions, and user bias, respectively. , , , , are the weighting coefficients corresponding to operating costs, carbon emissions, and user bias in the objective function, respectively; , These are the maximum and minimum operating costs obtained based on historical optimization results or simulations, respectively. , These are the maximum and minimum carbon emissions obtained based on historical optimization results or simulations, respectively. This represents the total number of users used to calculate user deviation.

[0012] Furthermore, the constraints include power balance constraints, charging station capacity constraints, power supply constraints, charging time window constraints, user adoption constraints, and upper and lower limits of state of charge constraints. The mathematical expression for the power balance constraint is as follows: In the formula, Indicates time period t The target charging power demand in the power grid planning; The mathematical expression for the capacity constraint of charging stations is as follows: In the formula, This indicates the rated capacity of the charging station; The mathematical expression for the power supply constraint is as follows: In the formula, Indicates that the power grid is t The maximum power that can be supplied to charge electric vehicles at any time; The mathematical expression for the charging time window constraint is as follows: In the formula, Indicates useru The earliest permitted start time for charging; Indicates the actual start time of charging; Indicates user u The latest permitted end time for charging; The mathematical expression for the user adoption constraint is as follows: In the formula, This represents the preset threshold for user adoption probability. ; The upper and lower limits of the state of charge are expressed mathematically as follows: In the formula, Indicates charging station s This corresponds to the lower limit of the state of charge of an electric vehicle battery. Indicates at time t At that time, it stopped at the charging station. s The actual state of charge of electric vehicle batteries; Indicates charging station s This corresponds to the upper limit of the state of charge of an electric vehicle battery.

[0013] Furthermore, based on the objective function and constraints, a joint optimization solution is performed to obtain the optimized charging scheme for each charging station, specifically including the following steps: Using power balance constraints as the coupling constraints of the system, Lagrange multipliers are introduced. The coupling constraints are relaxed and incorporated into the objective function to form a Lagrange dual function, the specific expression of which is as follows: In the formula, Represent the Lagrange dual function; N Indicates the total number of charging stations; Indicates that for the first s The Lagrangian function of an independent subproblem constructed from a charging station; The joint optimization problem consisting of the objective function and constraints is decomposed into N independent subproblems for charging stations. The Lagrangian function expressions for the independent subproblems are as follows: In the formula, Indicates the first s The decision variable vector of a charging station throughout the entire scheduling cycle; , , They represent the vectors of decision variables respectively. Calculate and normalize operating costs, carbon emissions, and user bias; Based on the Lagrange dual function and the Lagrange functions of each independent subproblem, the Lagrange dual decomposition algorithm is used for iterative solution. The following steps are performed during the iteration process: The independent subproblems of each charging station are solved in parallel to obtain the current iteration number. k The corresponding Lagrange multiplier and corresponding charging scheme vectors ; The Lagrange multipliers are updated using the subgradient method. The mathematical expression for the specific update process is as follows: In the formula, Indicates the step size; Indicate the Lagrange multiplier for the next iteration step; The Lagrange multipliers are iteratively updated until the preset convergence conditions are met, thus obtaining the optimized charging scheme for each charging station.

[0014] Furthermore, based on the management personnel's scores of the execution results, the weighting coefficients of carbon emissions, operating costs, and user deviations in the objective function for the next scheduling cycle are updated, specifically including the following steps: The weighted coefficients of carbon emissions, operating costs, and user bias together constitute the weighted parameter vector. , ; Based on the scoring feedback, metacognitive learning is used to update the weight parameter vector. The specific expression is as follows: In the formula, This represents the updated weight parameter vector; This represents the weight parameter vector before the update; This represents a vector of weight adjustments predetermined based on the scoring feedback; Indicates the learning rate; For the updated weight parameter vector The weighted parameters are normalized to obtain the weighted coefficients of carbon emissions, operating costs and user deviations in the objective function for the next scheduling cycle.

[0015] The beneficial effects of this invention are: This invention achieves comprehensive perception of the power supply, network, and charging load status by acquiring data from the power grid energy management system, distribution management system, and charging operation management platform. Based on the principle of economic dispatch, it identifies marginal generating units in the power grid at the current moment and calculates marginal carbon emission factors based on their carbon emission coefficients. This factor is then corrected using network-side data to obtain an effective marginal carbon emission factor, making the calculation of carbon emissions from electric vehicle charging behavior more accurate and significantly overcoming the biases of traditional calculations based on static or average factors. Simultaneously, it predicts the probability of users adopting charging schemes using a logistic regression model, further calculating user bias, enabling the charging dispatch process to effectively respond to users' actual intentions. Carbon emissions are normalized. By considering the quantity, operating costs, and user deviation, an objective function is constructed to minimize the weighted sum of these three indicators. This function is then used to collaboratively optimize the solution by considering system operation, charging facility capacity, and user demand constraints, thus achieving unified scheduling of city-level charging resources. Furthermore, by introducing a human-in-the-loop feedback mechanism, operation and management personnel can dynamically adjust the weight coefficients of carbon emissions, operating costs, and user deviation in the objective function based on the execution results of each scheduling cycle, forming a closed-loop optimization. This approach ensures low-carbon operation of the power grid while also considering economic efficiency and user satisfaction. It effectively solves the problems in existing technologies, such as insufficient accuracy in carbon emission accounting, neglect of spatial differences, and lack of real-time feedback and adjustment mechanisms, which make it difficult to achieve collaborative optimization of power resources and low-carbon operation of the power grid. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the overall process of the human-environment-electricity-carbon synergistic optimization method of the present invention; Figure 2 This is a schematic diagram of the overall interaction relationship between the source, network, load, and human in this invention; Figure 3 This is a diagram illustrating the electro-carbon coupling and economic dispatch mechanism of the present invention; Figure 4 This is a schematic diagram of the spatial network loss effect of the present invention; Figure 5 This is a schematic diagram of the human-in-the-loop feedback closed-loop process of the present invention; Figure 6 This is a schematic diagram of the Lagrange dual decomposition algorithm of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] like Figures 1 to 6 As shown, a method for co-optimizing electric vehicle charging resources in terms of human, environmental, and carbon emissions includes the following steps: Acquire multi-source heterogeneous data, including power supply side data, network side data, and load side data; Based on the principle of economic dispatch, marginal generating units in the power grid at the current moment are identified, and the marginal carbon emission factor of the power grid is calculated based on the carbon emission coefficient of the marginal generating units. The mechanism of power-carbon coupling and economic dispatch (Merit Order) is as follows: Figure 3 As shown; The marginal carbon emission factor is corrected using network-side data to obtain the effective marginal carbon emission factor, where the correction effect of network losses is considered. Figure 4 As shown; The carbon emissions of electric vehicle charging behavior are calculated based on the effective marginal carbon emission factor, and the operating costs are calculated based on load-side data. The probability of users adopting charging solutions is predicted by a logistic regression model to calculate user bias. The carbon emissions, operating costs, and user bias are normalized, and an objective function is constructed to minimize the weighted sum of carbon emissions, operating costs, and user bias. Constraints are established based on system operation, charging facility capacity, and user demand. Based on the objective function and constraints, a joint optimization solution is performed to obtain the optimal charging scheme for each charging station. After the optimized charging plan is executed in each scheduling cycle, the weighting coefficients of carbon emissions, operating costs, and user deviations in the objective function for the next scheduling cycle are updated based on the management personnel's score of the execution results. The human-loop feedback closed-loop process is as follows: Figure 5 As shown; This invention achieves comprehensive perception of the power supply, network, and charging load status by acquiring data from the power grid energy management system, distribution management system, and charging operation management platform. Based on the principle of economic dispatch, it identifies marginal generating units in the power grid at the current moment and calculates marginal carbon emission factors based on their carbon emission coefficients. This factor is then corrected using network-side data to obtain an effective marginal carbon emission factor, making the calculation of carbon emissions from electric vehicle charging behavior more accurate and significantly overcoming the biases of traditional calculations based on static or average factors. Simultaneously, it predicts the probability of users adopting charging schemes using a logistic regression model, further calculating user bias, enabling the charging dispatch process to effectively respond to users' actual intentions. Carbon emissions are normalized. By considering the quantity, operating costs, and user deviation, an objective function is constructed to minimize the weighted sum of these three indicators. This function is then used to collaboratively optimize the solution by considering system operation, charging facility capacity, and user demand constraints, thus achieving unified scheduling of city-level charging resources. Furthermore, by introducing a human-in-the-loop feedback mechanism, operation and management personnel can dynamically adjust the weight coefficients of carbon emissions, operating costs, and user deviation in the objective function based on the execution results of each scheduling cycle, forming a closed-loop optimization. This approach ensures low-carbon operation of the power grid while also considering economic efficiency and user satisfaction. It effectively solves the problems in existing technologies, such as insufficient accuracy in carbon emission accounting, neglect of spatial differences, and lack of real-time feedback and adjustment mechanisms, which make it difficult to achieve collaborative optimization of power resources and low-carbon operation of the power grid.

[0020] Data on the power supply side is obtained from the grid energy management system, data on the network side is obtained from the distribution management system, and data on the load side is obtained from the charging operation management platform. The overall system architecture is as follows: Figure 2 As shown; Power supply side data includes the real-time output power of each generator set, the upper and lower limits of the technical output of each generator set, the marginal operating cost quotation, and the carbon emission coefficient; Network-side data includes the power grid topology, voltage phasors at each node, and line resistance parameters; Load-side data includes the real-time charging power of each charging station, the state of charge of the batteries of each vehicle being charged, the estimated dwell time, and the user's historical charging behavior data.

[0021] The marginal carbon emission factor of the power grid is calculated based on the carbon emission coefficient of the marginal generating units. The specific calculation formula is as follows: in, Indicates the marginal carbon emission factor; Represents a set of generator sets; i Indicates generator set; Indicates generator set i The carbon emission coefficient; This indicates that the total load of the power grid is Under the operating conditions, the generator set i At the present moment t The output power is And only marginal units at the current moment t Output power It is a non-zero value.

[0022] The specific calculation formula for correcting the marginal carbon emission factor using network-side data is as follows: in, Indicates the effective marginal carbon emission factor; Indicates the marginal transmission loss rate; V Represents the node voltage phasor; R This indicates the line resistance parameter.

[0023] The carbon emissions from electric vehicle charging are calculated based on the effective marginal carbon emission factor, using the following formula: in, This represents the total carbon emissions generated by all scheduled electric vehicle charging activities. , This represents a charging scheme vector, indicating each charging station. s During the period t Power and location selection; The specific formula for calculating operating costs is as follows: in, Indicates operating costs; Indicates the charging price; The mathematical expression for the logistic regression model is as follows: in, The probability of users adopting a charging solution; The price discount represents the percentage difference between the actual electricity price in the optimized charging plan and the preset benchmark electricity price for that period. The carbon emission reduction margin represents the proportion of carbon emission reduction achieved by users charging according to the optimized charging plan compared to unordered charging. The convenience score is calculated based on the time and space convenience of the optimized charging solution. User bias is calculated based on the probability of users adopting charging solutions. The specific calculation formula is as follows: in, Indicates user bias; This indicates that, given a charging scheme vector Under these conditions, users u Probability of adoption of charging solution; Based on users' historical charging behavior data, the benchmark electricity price can be preset as the average electricity price for that period or the user's historical average payment price.

[0024] Construct an objective function that minimizes the weighted sum of carbon emissions, operating costs, and user bias, as shown in the following expression: in, Represent the objective function; , , These are the normalized operating costs, carbon emissions, and user bias, respectively. , , , , are the weighting coefficients corresponding to operating costs, carbon emissions, and user bias in the objective function, respectively; , These are the maximum and minimum operating costs obtained based on historical optimization results or simulations, respectively. , These are the maximum and minimum carbon emissions obtained based on historical optimization results or simulations, respectively. This represents the total number of users used to calculate user deviation.

[0025] The constraints include power balance constraints, charging station capacity constraints, power supply constraints, charging time window constraints, user adoption constraints, and upper and lower limits of state of charge constraints. The mathematical expression for the power balance constraint is as follows: In the formula, Indicates time period t The target charging power demand in the power grid planning; The mathematical expression for the capacity constraint of charging stations is as follows: In the formula, This indicates the rated capacity of the charging station; The mathematical expression for the power supply constraint is as follows: In the formula, Indicates that the power grid is t The maximum power that can be supplied to charge electric vehicles at any time; The mathematical expression for the charging time window constraint is as follows: In the formula, Indicates user u The earliest permitted start time for charging; Indicates the actual start time of charging; Indicates user u The latest permitted end time for charging; The mathematical expression for the user adoption constraint is as follows: In the formula, This represents the preset threshold for user adoption probability. ; Preferred, The value is set to 0.6, which means that on average, more than 60% of users are willing to adopt the optimization scheme in order to achieve a compromise between improving the efficiency of power grid operation and ensuring user experience. The upper and lower limits of the state of charge are expressed mathematically as follows: In the formula, Indicates charging station s The lower limit of the state of charge of the electric vehicle battery is the minimum percentage of charge allowed at the charging station. It is generally set according to the vehicle's range and battery protection requirements, such as 0.1 or 0.2, which corresponds to 10% or 20% of the charge. Indicates at time t At that time, it stopped at the charging station. s The actual state of charge of electric vehicle batteries; Indicates charging station s The upper limit of the state of charge of the electric vehicle battery is the maximum allowable charge ratio of the charging station, which is generally no more than 1. It can be set to 0.9 to 1.0 according to the battery health management requirements. For example, 0.9 means that the charge is controlled within 90% to slow down battery aging.

[0026] Based on the objective function and constraints, a joint optimization solution is performed to obtain the optimal charging scheme for each charging station. The specific steps include: Using power balance constraints as the coupling constraints of the system, Lagrange multipliers are introduced. The coupling constraints are relaxed and incorporated into the objective function to form a Lagrange dual function, the specific expression of which is as follows: In the formula, Represent the Lagrange dual function; N Indicates the total number of charging stations; Indicates that for the first s The Lagrangian function of each charging station's independent subproblem is used to describe the Lagrangian multipliers given a Lagrangian multiplier. μ In this case, the cost of the charging station under the combined effect of its local objective function and local constraints; The joint optimization problem consisting of the objective function and constraints is decomposed into: N The independent subproblems of each charging station are expressed by the Lagrange function as follows: In the formula, Indicates the first s The decision variable vector of a charging station throughout the entire scheduling cycle; , , They represent the vectors of decision variables respectively. Calculate and normalize operating costs, carbon emissions, and user bias; Based on the Lagrange dual function and the Lagrange functions of each independent subproblem, the Lagrange dual decomposition algorithm is adopted (the logic diagram of the Lagrange dual decomposition algorithm is shown in Figure 1). Figure 6 (As shown) Perform iterative solution, and execute the following steps during the iteration process: The independent subproblems of each charging station are solved in parallel to obtain the current iteration number. k The corresponding Lagrange multiplier and corresponding charging scheme vectors ; The Lagrange multipliers are updated using the subgradient method. The mathematical expression for the specific update process is as follows: In the formula, Indicates the step size; Indicate the Lagrange multiplier for the next iteration step; Iteratively update the Lagrange multipliers until the preset convergence conditions are met, and obtain the optimized charging scheme for each charging station. This represents the violation of power balance constraints. The preset convergence condition is that the absolute value of the violation is less than a preset convergence threshold. The specific mathematical expression is as follows: Preferred convergence threshold It can be set to 1% to 2% of the target power requirement; For example, in a simulation system for a megacity, the target charging power demand during a typical time period is... Approximately 100MW. To ensure the power balance constraint has an acceptable accuracy in engineering, a preset convergence threshold is set. 1% of the target charging power demand (i.e., 1MW).

[0027] Based on the management personnel's scores of the execution results, update the weighting coefficients of carbon emissions, operating costs, and user deviations in the objective function for the next scheduling cycle. This includes the following steps: The weighted coefficients of carbon emissions, operating costs, and user bias together constitute the weighted parameter vector. , ; Based on the scoring feedback, metacognitive learning is used to update the weight parameter vector. The specific expression is as follows: In the formula, This represents the updated weight parameter vector; This represents the weight parameter vector before the update; This represents a vector of pre-determined weight adjustments based on the scoring feedback, used to characterize the magnitude by which each weight needs to be increased or decreased. R This indicates the evaluation results of operations and management personnel. Indicates system operating performance indicators, e Indicates evaluation error; Indicates the learning rate; For the updated weight parameter vector The weighted parameters are normalized to obtain the weighted coefficients of carbon emissions, operating costs and user deviations in the objective function for the next scheduling cycle.

[0028] To verify the effectiveness of this invention, simulation verification was performed based on typical power grid data from a megacity: Power grid scale: It includes 10 major power plants (coal, gas, hydro, wind and solar), and the distribution network includes 118 nodes.

[0029] Charging facilities: Simulates 500 public charging stations, distributed in commercial, residential and industrial areas.

[0030] Vehicle scale: Simulates the daily charging demand of 50,000 electric vehicles.

[0031] Comparison with benchmark: Baseline 1 (Disorderly Charging): Users charge as soon as they arrive, without considering electricity prices or carbon emissions; Baseline 2 (Traditional Ordered Charging): Optimized solely based on Time-of-Use (TOU) pricing and using the annual average carbon emission factor; Proposed (this invention): Based on dynamic MCEF, space loss correction, and human-in-the-loop feedback; Table 1. Performance indicators for different charging methods While traditional orderly charging (Baseline 2) utilizes low electricity prices to shift load to nighttime, its carbon reduction effect is limited due to potential peak shaving by thermal power plants at night. This invention, however, accurately identifies "wind curtailment" times and "low-loss nodes," achieving true deep decarbonization. Particularly in spatial optimization, this invention successfully guides approximately 15% of the load from remote high-loss sites to hub low-loss sites, significantly reducing system-level energy consumption.

[0032] The simulation covered a 30-day operation process, with the initial weights set to equal.

[0033] Days 1-5: The system was in the exploratory phase. Administrators frequently reported "excessive costs" or "user complaints about inconvenient charging." The MCL algorithm was quickly adjusted, and additional measures were taken. , ; Days 10-20: The system enters a stable period, and weight fluctuations decrease; this coincides with a "severe air pollution warning," and the administrator adjusts the feedback strategy, emphasizing low carbon emissions; the system responds quickly and improves performance. Although this resulted in a slight increase in costs, it significantly reduced carbon emissions.

[0034] Day 30: The user adoption rate stabilized at around 78% (a huge improvement compared to the traditional 40-50%), proving the key role of the "human-in-the-loop" mechanism in improving the system's usability.

[0035] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0036] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the present invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention.

Claims

1. A method for synergistic optimization of human, environmental, and carbon emissions in electric vehicle charging resources, characterized in that, Includes the following steps: Acquire multi-source heterogeneous data, including power supply side data, network side data, and load side data; Based on the principle of economic dispatch, marginal units in the power grid at the current moment are identified, and the marginal carbon emission factor of the power grid is calculated based on the carbon emission coefficient of the marginal units. By using network-side data to correct the marginal carbon emission factor, an effective marginal carbon emission factor can be obtained. The carbon emissions of electric vehicle charging behavior are calculated based on the effective marginal carbon emission factor, and the operating costs are calculated based on load-side data. The probability of users adopting charging solutions is predicted by a logistic regression model to calculate user bias. The carbon emissions, operating costs, and user bias are normalized, and an objective function is constructed to minimize the weighted sum of carbon emissions, operating costs, and user bias. Constraints are established based on system operation, charging facility capacity, and user demand. Based on the objective function and constraints, a joint optimization solution is performed to obtain the optimal charging scheme for each charging station. After the optimized charging plan is executed in each scheduling cycle, the weighting coefficients of carbon emissions, operating costs and user deviation in the objective function for the next scheduling cycle are updated based on the management personnel's score of the execution results.

2. The method for synergistic optimization of human-environmental-carbon energy charging resources for electric vehicles according to claim 1, characterized in that, Power supply side data includes the real-time output power of each generator set, the upper and lower limits of the technical output of each generator set, the marginal operating cost quotation, and the carbon emission coefficient; Network-side data includes the power grid topology, voltage phasors at each node, and line resistance parameters; Load-side data includes the real-time charging power of each charging station, the state of charge of the batteries of each vehicle being charged, the estimated dwell time, and the user's historical charging behavior data.

3. The method for synergistic optimization of human-environmental-carbon aspects of electric vehicle charging resources according to claim 2, characterized in that, The marginal carbon emission factor of the power grid is calculated based on the carbon emission coefficient of the marginal generating units. The specific calculation formula is as follows: in, Indicates the marginal carbon emission factor; Represents a set of generator sets; i Indicates generator set; Indicates generator set i The carbon emission coefficient; This indicates that the total load of the power grid is Under the operating conditions, the generator set i At the present moment t The output power is And only marginal units at the current moment t Output power It is a non-zero value.

4. The method for synergistic optimization of human-environmental-carbon energy charging resources for electric vehicles according to claim 3, characterized in that, The specific calculation formula for correcting the marginal carbon emission factor using network-side data is as follows: in, Indicates the effective marginal carbon emission factor; Indicates the marginal transmission loss rate; V Represents the node voltage phasor; R This indicates the line resistance parameter.

5. The method for coordinated optimization of human, environmental, and carbon aspects of electric vehicle charging resources according to claim 4, characterized in that, The carbon emissions from electric vehicle charging are calculated based on the effective marginal carbon emission factor, using the following formula: in, This represents the total carbon emissions generated by all scheduled electric vehicle charging activities. , This represents a charging scheme vector, indicating each charging station. s During the period t Power and location selection; The specific formula for calculating operating costs is as follows: in, Indicates operating costs; Indicates the charging price; The mathematical expression for the logistic regression model is as follows: in, The probability of users adopting a charging solution; The price discount represents the percentage difference between the actual electricity price in the optimized charging plan and the preset benchmark electricity price for that period. The carbon emission reduction margin represents the proportion of carbon emission reduction achieved by users charging according to the optimized charging plan compared to unordered charging. The convenience score is calculated based on the time and space convenience of the optimized charging solution. User bias is calculated based on the probability of users adopting charging solutions. The specific calculation formula is as follows: in, Indicates user bias; This indicates that, given a charging scheme vector Under these conditions, users u Probability of adopting a charging solution.

6. The method for coordinated optimization of human, environmental, and carbon aspects of electric vehicle charging resources according to claim 5, characterized in that, Construct an objective function that minimizes the weighted sum of carbon emissions, operating costs, and user bias, as shown in the following expression: in, Represent the objective function; , , These are the normalized operating costs, carbon emissions, and user bias, respectively. , , , , are the weighting coefficients corresponding to operating costs, carbon emissions, and user bias in the objective function, respectively; , These are the maximum and minimum operating costs obtained based on historical optimization results or simulations, respectively. , These are the maximum and minimum carbon emissions obtained based on historical optimization results or simulations, respectively. This represents the total number of users used to calculate user deviation.

7. The method for synergistic optimization of human-environmental-carbon energy charging resources for electric vehicles according to claim 6, characterized in that, The constraints include power balance constraints, charging station capacity constraints, power supply constraints, charging time window constraints, user adoption constraints, and upper and lower limits of state of charge constraints. The mathematical expression for the power balance constraint is as follows: In the formula, Indicates time period t The target charging power demand in the power grid planning; The mathematical expression for the capacity constraint of charging stations is as follows: In the formula, This indicates the rated capacity of the charging station; The mathematical expression for the power supply constraint is as follows: In the formula, Indicates that the power grid is t The maximum power that can be supplied to charge electric vehicles at any time; The mathematical expression for the charging time window constraint is as follows: In the formula, Indicates user u The earliest permitted start time for charging; Indicates the actual start time of charging; Indicates user u The latest permitted end time for charging; The mathematical expression for the user adoption constraint is as follows: In the formula, This represents the preset threshold for user adoption probability. ; The upper and lower limits of the state of charge are expressed mathematically as follows: In the formula, Indicates charging station s This corresponds to the lower limit of the state of charge of an electric vehicle battery. Indicates at time t At that time, it stopped at the charging station. s The actual state of charge of electric vehicle batteries; Indicates charging station s This corresponds to the upper limit of the state of charge of an electric vehicle battery.

8. The method for synergistic optimization of human-environmental-carbon energy charging resources for electric vehicles according to claim 7, characterized in that, Based on the objective function and constraints, a joint optimization solution is performed to obtain the optimal charging scheme for each charging station. The specific steps include: Using power balance constraints as the coupling constraints of the system, Lagrange multipliers are introduced. The coupling constraints are relaxed and incorporated into the objective function to form a Lagrange dual function, the specific expression of which is as follows: In the formula, Represent the Lagrange dual function; N Indicates the total number of charging stations; Indicates that for the first s The Lagrangian function of an independent subproblem constructed from a charging station; The joint optimization problem consisting of the objective function and constraints is decomposed into: N The independent subproblems of each charging station are expressed by the Lagrange function as follows: In the formula, Indicates the first s The decision variable vector of a charging station throughout the entire scheduling cycle; , , They represent the vectors of decision variables respectively. Calculate and normalize operating costs, carbon emissions, and user bias; Based on the Lagrange dual function and the Lagrange functions of each independent subproblem, the Lagrange dual decomposition algorithm is used for iterative solution. The following steps are performed during the iteration process: The independent subproblems of each charging station are solved in parallel to obtain the current iteration number. k The corresponding Lagrange multiplier and corresponding charging scheme vectors ; The Lagrange multipliers are updated using the subgradient method. The mathematical expression for the specific update process is as follows: In the formula, Indicates the step size; Indicate the Lagrange multiplier for the next iteration step; The Lagrange multipliers are iteratively updated until the preset convergence conditions are met, thus obtaining the optimized charging scheme for each charging station.

9. The method for coordinated optimization of human, environmental, and carbon aspects of electric vehicle charging resources according to claim 8, characterized in that, Based on the management personnel's scores of the execution results, update the weighting coefficients of carbon emissions, operating costs, and user deviations in the objective function for the next scheduling cycle. This includes the following steps: The weighted coefficients of carbon emissions, operating costs, and user bias together constitute the weighted parameter vector. , ; Based on the scoring feedback, metacognitive learning is used to update the weight parameter vector. The specific expression is as follows: In the formula, This represents the updated weight parameter vector; This represents the weight parameter vector before the update; This represents a vector of weight adjustments predetermined based on the scoring feedback; Indicates the learning rate; For the updated weight parameter vector The weighted parameters are normalized to obtain the weighted coefficients of carbon emissions, operating costs and user deviations in the objective function for the next scheduling cycle.

Images