Distributed flexible load regulation and control method and system for building group and electric vehicle

Abstract

The invention discloses a distributed flexible load regulation and control method and system for a building group and an electric vehicle, and the method comprises the steps: employing a recursive state space model to predict the air conditioning load and illumination load of the building group and the charging load of the electric vehicle according to the energy consumption data of the building group, the environment data and the charging behavior data of the electric vehicle; then, an optimization function with peak load shifting, electric charge reduction and user demand guarantee as targets is established, and a joint scheduling scheme of distributed flexible load regulation and control is solved by adopting a tabu search algorithm under the constraint conditions of the comfort level of the building group, the SOC demand of the electric vehicle and the total load of the power grid; and finally, issuing the joint scheduling scheme to a building group controller and a charging pile controller, and executing distributed flexible load regulation and control. According to the invention, combined prediction and regulation of the air conditioner and lighting load of the building group and the charging load of the electric vehicle are realized, and the stability of power grid operation is enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of flexible load management and distributed energy regulation of power systems, in particular to a distributed flexible load regulation method and system for building groups and electric vehicles. BACKGROUND

[0002] With the increasing penetration of renewable energy (such as photovoltaic and wind power) in distribution networks, the operation of power systems presents stronger volatility and uncertainty, which brings challenges to the safe and stable operation of power grids. At the same time, the proportion of building group energy consumption (especially air conditioning and lighting) in urban power load is gradually increasing, and the large-scale access of electric vehicles also makes the distribution network face greater load pressure during the charging peak period. Under this background, how to utilize the air conditioning and lighting loads with adjustment flexibility in building groups and the transferability of electric vehicle charging to achieve the coordinated regulation of multi-source flexible loads has become a key problem to improve the stability and energy efficiency of power grids.

[0003] Existing flexible load regulation methods mainly include centralized optimization and local independent control. The centralized optimization method needs to collect detailed operation data of all buildings and electric vehicles, which has great calculation and communication overhead, and is difficult to apply in real time in large-scale scenarios, and there is a serious risk of user privacy leakage. In the local independent control method, each building or charging pile only focuses on its own load optimization, lacks overall coordination with the grid target, and is easy to lead to local optimization rather than global optimization, which has limited peak clipping and valley filling effect. SUMMARY

[0004] The purpose of the present application is to provide a distributed flexible load regulation method and system for building groups and electric vehicles, which can effectively solve the problem of charging load uncertainty caused by the large-scale access of electric vehicles to building groups and the power grid fluctuation caused by the superposition of building flexible loads.

[0005] The technical scheme of the present application is a distributed flexible load regulation method for building groups and electric vehicles, characterized in that it comprises the following steps:

[0006] Collecting building group energy consumption data, environmental data and electric vehicle charging behavior data;

[0007] According to the building group energy consumption data, environmental data and electric vehicle charging behavior data, a recursive state space model is used to predict the air conditioning load and lighting load of the building group, and the charging load of the electric vehicle;

[0008] An optimization function is established to achieve the goals of peak clipping and valley filling, reducing electricity bills and protecting user demand, and a tabu search algorithm is used to solve the joint scheduling scheme of distributed flexible load regulation under the constraints of building group comfort, electric vehicle SOC demand and total grid load.

[0009] The joint scheduling scheme is issued to the building cluster controller and the charging pile controller to perform distributed flexible load regulation.

[0010] Further, a rolling horizon optimization mechanism is adopted for dynamic distributed flexible load regulation. In each scheduling period, building cluster energy consumption data, environmental data, and electric vehicle charging behavior data are collected, a joint scheduling scheme for distributed flexible load regulation is solved, and the joint scheduling scheme is issued to the building cluster controller and the charging pile controller to perform distributed flexible load regulation.

[0011] Further, according to the building cluster energy consumption data, environmental data, and electric vehicle charging behavior data, a recursive state space model is used to predict the air conditioning load and lighting load of the building cluster, and the charging load of the electric vehicle, including:

[0012] For the air conditioning system, a first-order thermodynamic RC model is used to describe the relationship between the building room temperature and the air conditioning power;

[0013] For the lighting system, the natural light compensation effect is considered to calculate the lighting power;

[0014] The implicit state vector is used to represent the potential dynamic characteristics of building energy consumption, and the state is recursively updated and the future air conditioning load and lighting load are generated and predicted in combination with the building external input information;

[0015] The building external input information includes outdoor temperature, light intensity, time characteristics, and / or electricity price signal.

[0016] Further, according to the building cluster energy consumption data, environmental data, and electric vehicle charging behavior data, a recursive state space model is used to predict the air conditioning load and lighting load of the building cluster, and the charging load of the electric vehicle, including:

[0017] For the electric vehicle charging system, an energy balance equation is used to describe the battery energy state;

[0018] The implicit state vector is used to represent the potential dynamic characteristics of electric vehicle charging demand, and the state is recursively updated and the future charging load is generated and predicted in combination with the vehicle external input information;

[0019] The vehicle external input information includes vehicle arrival time, departure time, initial SOC, battery capacity, and / or driving distance.

[0020] Further, an optimization function is established to achieve the goals of peak load shifting, electricity cost reduction, and user demand guarantee, including:

[0021] The optimization function is to minimize the weighted sum of peak load, electricity cost, comfort loss, and electric vehicle charging penalty.

[0022] Further, under the constraints of building group comfort, electric vehicle SOC demand and total grid load, the joint scheduling scheme of distributed flexible load regulation is solved by using tabu search algorithm, including:

[0023] The total grid load constraint is that the sum of air conditioning load, lighting load and charging load is not greater than the upper limit of the capacity of the distribution network.

[0024] Further, the building group energy consumption data, environmental data and electric vehicle charging behavior data are collected, including:

[0025] The building group energy consumption data includes air conditioning power, lighting power, indoor temperature, humidity and illuminance data of each unit in the building group;

[0026] The environmental data includes outdoor temperature, light intensity and weather characteristics;

[0027] The electric vehicle charging behavior data includes the arrival time, departure time, real-time state of charge, rated capacity and driving distance of the vehicle.

[0028] The distributed flexible load regulation system for building group and electric vehicle according to the application comprises:

[0029] The data acquisition module is used to collect building group energy consumption data, environmental data and electric vehicle charging behavior data;

[0030] The load prediction module is used to predict the air conditioning load and lighting load of the building group and the charging load of the electric vehicle according to the building group energy consumption data, environmental data and electric vehicle charging behavior data by using the recursive state space model;

[0031] The optimization control module is used to establish an optimization function with the goals of peak shaving, reducing electricity bills and guaranteeing user demand, and under the constraints of building group comfort, electric vehicle SOC demand and total grid load, the joint scheduling scheme of distributed flexible load regulation is solved by using tabu search algorithm;

[0032] The execution module is used to issue the joint scheduling scheme to the building group controller and the charging pile controller to execute distributed flexible load regulation.

[0033] The electronic device according to the application comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the computer program realizes the distributed flexible load regulation method for building group and electric vehicle when loaded into the processor.

[0034] The computer readable storage medium of the application stores a computer program, and the computer program is executed by a processor to realize the distributed flexible load regulation method for building groups and electric vehicles.

[0035] Advantages: Compared with the prior art, the application has the advantages that: the application combines a deep learning prediction model (RSSM) and an optimization algorithm (tabu search) to realize joint prediction and regulation of building air conditioning and lighting loads and electric vehicle charging loads. The application considers uncertain factors such as arrival time, departure time, battery capacity, initial SOC and driving distance of electric vehicles in load prediction, and dynamically models the energy consumption law of building groups; in the optimization and regulation process, the total energy consumption of the building is not over limited, the user comfort and the electric vehicle charging completion rate are constrained, and a multi-level regulation architecture is used to realize the collaborative optimization of building load and electric vehicle load. Compared with the prior art, the application can significantly improve the accuracy of load prediction through RSSM, effectively cope with the uncertainty brought by the access of electric vehicles; on the building side, flexible regulation is realized through pre-cooling, temperature set point fine-tuning and natural light compensation, on the electric vehicle side, charging strategy is dynamically adjusted combined with time-of-use electricity price and photovoltaic output, the flexibility and reliability of the overall regulation are improved; at the same time, the distributed optimization architecture is adopted to ensure the scalability and privacy protection of the system, and the safe operation of the power grid is realized under the global coordination, so as to effectively peak shaving and valley filling, reduce the electricity cost, improve the renewable energy consumption level and enhance the stability of the power distribution network operation, on the premise of meeting the user comfort and travel demand. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 The distributed flexible load regulation method of the embodiment of the application is a flowchart.

[0037] Figure 2 The distributed flexible load regulation method of the embodiment of the application is a framework diagram.

[0038] Figure 3 The distributed flexible load regulation method of the embodiment of the application is a system architecture diagram DETAILED DESCRIPTION

[0039] The present application improves the accuracy and robustness of load forecasting by introducing RSSM to dynamically model the air conditioning and lighting loads of the building group and the electric vehicle charging load, fully considering the uncertainty factors such as arrival time, departure time, SOC, battery capacity and driving distance, etc. Further, the tabu search algorithm is used to jointly optimize the building flexible load and electric vehicle charging strategy, effectively avoiding the problem that the traditional optimization method is easy to fall into local optimum, realizing the comprehensive regulation and control of air conditioning pre-cooling, lighting natural light compensation and electric vehicle charging power distribution under the condition that the total load of the building does not exceed the limit, significantly improving the peak clipping and valley filling effect, reducing the electricity cost of users, and improving the utilization rate of renewable energy. Finally, through the interaction of the local controller and the central coordinator, distributed dynamic regulation and control is realized, which not only guarantees the user comfort and electric vehicle travel demand, but also enhances the stability of the power system in the high penetration rate scenario. The technical solutions of the present application are further described below in combination with the drawings.

[0040] As Figure 1 shown is the flow chart of the distributed flexible load regulation and control method of the present application, as Figure 2 shown is the framework diagram of the distributed flexible load regulation and control method of the present application, in combination with Figure 1 and Figure 2 , the distributed flexible load regulation and control method of the present application comprises the following steps.

[0041] Step S1, data acquisition step: acquiring building group energy consumption data, environmental data and electric vehicle charging behavior data;

[0042] Step S11, building group data: through the building energy management system (BEMS) and the Internet of Things sensor network, real-time acquisition of air conditioning power, lighting power, indoor temperature, humidity and illuminance data of each unit in the building group;

[0043] Step S12, electric vehicle data: through the communication interface of the electric vehicle charging pile, the arrival time departure time real-time state of charge of the battery battery rated capacity Cap j , and driving distance d j of the vehicle are acquired;

[0044] Step S13, environmental data: through the application program interface (API), the weather forecast data of the next 24 hours are acquired, including outdoor temperature, light intensity and weather characteristics;

[0045] Step S14, the collected multi-source heterogeneous data are cleaned, aligned and normalized to form a model input data set.

[0046] Step S2, load prediction step: recursive state space model (RSSM) is used to predict the building cluster air conditioning and lighting load and electric vehicle charging load;

[0047] Step S21, building cluster load prediction:

[0048] The historical energy consumption data is preprocessed and feature extracted;

[0049] For the air conditioning system, a first-order thermodynamic RC model is used to describe the relationship between the building room temperature and the air conditioning power:

[0050]

[0051] where C i is the heat capacity of building unit i, R i is the thermal resistance of building unit i, T i (t) is the indoor temperature at time t, T out (t) is the outdoor temperature at time t, η i is the air conditioning system efficiency coefficient, is the air conditioning power of building unit i at time t, is the heat generated by the building internal personnel, equipment, etc.

[0052] For the lighting system, the natural light compensation effect is considered:

[0053]

[0054] where, is the lighting power of building unit i at time t, is the baseline lighting load without natural light, κ i is the natural light compensation coefficient, L nat (t) is the natural light intensity at time t, ∈ i (t) is noise or unmodeled error, which needs to satisfy the constraint

[0055] On this basis, the recursive state space model (RSSM) is used to predict the building cluster air conditioning and lighting load:

[0056] The state transition formula is:

[0057] s t ~p θ (s t |s t-1 ,u t-1 ,a t-1 )

[0058] where s t is the hidden state vector of the system, u t-1is the previous control input, a t-1 is the external input, and Θ is the RSSM model parameters.

[0059] The observation generation formula is:

[0060]

[0061] where s t is the current state, d t is the current external disturbance, is the predicted air conditioning load power at time t, is the predicted lighting load power at time t.

[0062] The total building cluster air conditioning load and lighting load prediction is:

[0063]

[0064] where, is the total air conditioning load of the building cluster, is the total lighting load of the building cluster.

[0065] Step S22, electric vehicle load prediction:

[0066] The electric vehicle battery energy state is described by the following energy balance equation:

[0067]

[0068] where E j (t) is the battery energy of electric vehicle j at time t, is the charging power of electric vehicle j at time t, is the discharging power of electric vehicle j at time t, satisfying η c is the charging efficiency, η d is the discharging efficiency, Δ t is the time step.

[0069] The constraint conditions include electric vehicle battery capacity constraints and SOC constraints:

[0070] Electric vehicle battery capacity constraint:

[0071] 0≤E j (t)≤Cap j ;

[0072] where Cap j is the battery capacity;

[0073] Electric vehicle departure SOC constraint:

[0074]

[0075] where, is the battery state of charge of the electric vehicle j at time t, is the vehicle departure time, is the minimum SOC requirement at vehicle departure.

[0076] When predicting the charging load, a recursive state space model (RSSM) is used to generate a probabilistic charging demand scenario; it considers input uncertainty parameters and predicts future power distribution through the collaborative work of an encoder, a recurrent state transition model, and a decoder:

[0077]

[0078] The expected charging power of each vehicle at each time is calculated, and then the expected values of all vehicles are added to obtain the benchmark prediction curve of the whole network. The overall electric vehicle load prediction is:

[0079]

[0080] where, is the total load of the electric vehicle group at time t, is the expected value of the charging power of vehicle j at time t.

[0081] Step S3, optimization and control step: an optimization function is established to achieve the goals of peak clipping and valley filling, reducing electricity bills, and ensuring user demand. Under the constraints of building comfort, electric vehicle SOC demand, and total grid load, a Tabu Search algorithm is used to solve the joint scheduling scheme.

[0082] Step S31, a Tabu Search (Tabu Search) algorithm is used to solve the optimal control strategy by minimizing the total electricity cost, load variance, and maximizing the photovoltaic consumption rate; an optimization objective function is established:

[0083] min F = α·P peak + β·C cost + γ·D comfort + δ·V violation ;

[0084]

[0085] where, P peak is the peak load, C cost is the electricity cost, c t is the time-of-use electricity price, D comfort is the comfort loss, V violation is the electric vehicle charging penalty.

[0086] The total grid load constraint needs to be met:

[0087]

[0088] P grid (t) is the total load of the power grid, is (including air conditioning load and lighting load), P max is the upper limit of the capacity of the power distribution network;

[0089] A global search is performed under the above constraints using a tabu search algorithm, and the neighborhood operations include: charging period adjustment, power distribution adjustment, air conditioning set temperature adjustment, and lighting power adjustment.

[0090] In step S32, the tabu search algorithm uses a tabu table mechanism to record the last several scheduling operations to avoid the search from falling into a local cycle; and the current optimal solution and the global optimal solution are retained during the iteration until the upper limit of the number of iterations or the convergence condition is met.

[0091] In step S4, a feedback step is performed: the optimized scheduling scheme is issued to the building group controller and the charging pile controller, and dynamic updating is performed based on the rolling time domain optimization mechanism.

[0092] In step S41, the central coordinator issues the generated control instruction set to the corresponding building automation system (BAS) field controller and intelligent charging pile controller, and issues the control instructions to the air conditioning, lighting system and electric vehicle charging pile through the execution module to realize the joint optimization control of flexible load and electric vehicle charging load.

[0093] In step S42, each local controller executes the instructions and feeds back the actual running state data of the system to the central coordinator.

[0094] In the air conditioning system, peak shifting is achieved through pre-cooling, delayed start and temperature set point fine adjustment;

[0095] In the lighting system, natural light compensation is performed in combination with the ambient illuminance sensor data to reduce the lighting power during the peak electricity consumption period.

[0096] On the electric vehicle charging pile side, based on the time-of-use electricity price and the photovoltaic output, the charging time and charging power are dynamically adjusted to achieve peak clipping and valley filling.

[0097] In step S43, the central coordinator starts rolling optimization based on the deviation between the actual and predicted values in the model predictive control (MPC) framework, re-executes the prediction and optimization process; and in each scheduling period t, the load in the next H time periods is predicted and optimized based on the prediction window H, and only the first step control decision is executed to ensure the dynamic adaptability of the system.

[0098] The distributed flexible load regulation system for the building group and the electric vehicle comprises a data acquisition module, a load prediction module, an optimization regulation module, an execution module and a central coordinator.

[0099] The data acquisition module is used for acquiring real-time data of the building group and the electric vehicle, including building group energy consumption data, environment data and electric vehicle charging behavior data.

[0100] The load prediction module is used for predicting air conditioning load and lighting load of the building group and charging load of the electric vehicle according to the building group energy consumption data, the environment data and the electric vehicle charging behavior data by using a recursive state space model.

[0101] The optimization regulation module is used for establishing an optimization function with the target of peak load shifting, reducing electricity bills and guaranteeing user demand, and solving a joint scheduling scheme of the distributed flexible load regulation by using a tabu search algorithm under the constraint conditions of building group comfort, electric vehicle SOC demand and total grid load.

[0102] The execution module is used for issuing the joint scheduling scheme to the building group controller and the charging pile controller to execute the distributed flexible load regulation.

[0103] The central coordinator is used for integrating and managing the modules and scheduling the whole process.

[0104] The modules in the load regulation system can be realized by software, hardware or a combination of software and hardware. The modules can be embedded into a processor of a computing device in a hardware form or exist as independent components, or be stored in a storage medium of the computing device in a software form, and be called and executed by the processor to realize corresponding functional operations.

[0105] The electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the computer program realizes the distributed flexible load regulation method for the building group and the electric vehicle when being loaded to the processor.

[0106] The computer readable storage medium stores a computer program, and the computer program realizes the distributed flexible load regulation method for the building group and the electric vehicle when being executed by the processor.

[0107] The computer readable storage medium can include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, flash memory or any other medium that can be used to store program codes in the form of instructions or data structures and can be accessed by a computer.

[0108] The processor is configured to execute the computer program stored in the memory to implement the various steps in the methods involved in the above-described embodiments.

Claims

1. A method for distributed flexible load regulation for a building cluster and electric vehicles, characterized in that, The method comprises the following steps: collecting building group energy consumption data, environmental data, and electric vehicle charging behavior data; predicting air conditioning load and lighting load of the building group and charging load of the electric vehicle according to the building group energy consumption data, environmental data, and electric vehicle charging behavior data by using a recursive state space model; establishing an optimization function with the goal of peak load shifting, reducing electricity bills, and guaranteeing user demand, and solving a joint scheduling scheme of distributed flexible load regulation under the constraints of building group comfort, electric vehicle SOC demand, and total grid load by using a tabu search algorithm; issuing the joint scheduling scheme to building group controllers and charging pile controllers to execute distributed flexible load regulation. 2.The method of claim 1, wherein, The method comprises the following steps: 3.The method of claim 1, wherein, collecting building group energy consumption data, environmental data, and electric vehicle charging behavior data in each scheduling period to solve a joint scheduling scheme of distributed flexible load regulation, and issuing the joint scheduling scheme to building group controllers and charging pile controllers to execute distributed flexible load regulation by using a rolling horizon optimization mechanism. The method comprises the following steps: for an air conditioning system, a first-order thermodynamic RC model is used to describe the relationship between building room temperature and air conditioning power; for a lighting system, lighting power is calculated by considering the natural light compensation effect; a hidden state vector is used to represent the potential dynamic characteristics of building energy consumption, and the state is recursively updated and future air conditioning load and lighting load are generated and predicted by combining external input information of the building; 4.The method of claim 1, wherein, the external input information of the building comprises outdoor temperature, light intensity, time characteristics, and / or electricity price signals. The method comprises the following steps: for an electric vehicle charging system, an energy balance equation is used to describe the battery energy state; a hidden state vector is used to represent the potential dynamic characteristics of electric vehicle charging demand, and the state is recursively updated and future charging load is generated and predicted by combining external input information of the vehicle; 5. The method of claim 1, wherein the method is characterized by, the external input information of the vehicle comprises vehicle arrival time, departure time, initial SOC, battery capacity, and / or driving distance. The method comprises the following steps: 6.The method of claim 1, wherein, the optimization function is to minimize the weighted sum of peak load, electricity bill cost, comfort loss, and electric vehicle charging penalty. The method comprises the following steps: 7.The method of claim 1, wherein, the total grid load constraint is that the sum of air conditioning load, lighting load, and charging load is not greater than the upper limit of the capacity of the distribution network. The method comprises the following steps: the building group energy consumption data comprises air conditioning power, lighting power, indoor temperature, humidity, and illuminance data of each unit in the building group; The environmental data includes outdoor temperature, light intensity and weather characteristics; The electric vehicle charging behavior data includes arrival time, departure time, real-time state of charge, rated capacity and driving distance of the vehicle.

8. A distributed flexible load regulation system for a building complex and electric vehicles, characterized by, The method comprises the steps of: a data collection module for collecting building group energy consumption data, environmental data and electric vehicle charging behavior data; a load prediction module for predicting air conditioning load and lighting load of the building group and charging load of the electric vehicle according to the building group energy consumption data, environmental data and electric vehicle charging behavior data by using a recursive state space model; an optimization control module for establishing an optimization function with the goal of peak shaving, reducing electricity bills and guaranteeing user demand, and solving a joint scheduling scheme of distributed flexible load regulation by using a tabu search algorithm under the constraints of building group comfort, electric vehicle SOC demand and total grid load; an execution module for issuing the joint scheduling scheme to building group controllers and charging pile controllers to execute distributed flexible load regulation.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The computer program is loaded into the processor to implement the distributed flexible load regulation method for building groups and electric vehicles according to any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program is executed by the processor to implement the distributed flexible load regulation method for building groups and electric vehicles according to any one of claims 1-7.

Images