CN-121984000-A - Charging and discharging control method and related device for elevator energy storage device

CN121984000ACN 121984000 ACN121984000 ACN 121984000ACN-121984000-A

Abstract

The application discloses a charge and discharge control method of an elevator energy storage device and a related device, relating to the technical field of elevator energy saving control, comprising the following steps: according to the electricity price prediction sequence and the elevator power prediction sequence in the prediction period, an action sequence optimization model comprising an action sequence, an objective function and constraint conditions is constructed, wherein the objective function represents the total net benefit of a system after actions in the action sequence are sequentially executed on the elevator energy storage device in the prediction period, the total net benefit is jointly determined by electricity charge benefit, the aging cost of the energy storage device and the abrasion cost of an energy storage device control component, when the objective function is solved with the maximum total purification benefit as the objective function, the aging of the energy storage device and the abrasion of the energy storage device control component are quantized into cost, and the cost and the electricity charge benefit are uniformly optimized, so that frequent operation of the energy storage device control component for pursuing tiny electricity charge benefit can be effectively avoided, or charge and discharge actions damaging the health of the energy storage device are carried out.

Inventors

HE TIANYU
Han Kunfeng
CHEN WEI

Assignees

合肥华思系统股份有限公司

Dates

Publication Date: 20260505
Application Date: 20260126

Claims (11)

1. The charge and discharge control method of the elevator energy storage device is characterized by comprising the following steps of: acquiring an electricity price prediction sequence and an elevator power prediction sequence in a prediction period; Constructing an action sequence optimization model according to the electricity price prediction sequence and the elevator power prediction sequence, wherein the action sequence optimization model comprises an action sequence, an objective function and constraint conditions, the objective function represents total net benefit of a system after the elevator energy storage device is sequentially executed with actions in the action sequence in the prediction period, and the total net benefit is determined by electricity charge benefit, energy storage device aging cost and energy storage device control component wear cost; with the total net benefit maximization as a target, solving the objective function by adopting a two-stage heuristic algorithm under the constraint of the constraint condition to obtain an optimal action sequence; And controlling the charge and discharge of the energy storage device in the prediction period according to the optimal action sequence.
2. The charge/discharge control method of an elevator energy storage device according to claim 1, wherein the acquiring the electricity price prediction sequence and the elevator power prediction sequence in the prediction period includes: acquiring a multidimensional feature matrix corresponding to a power grid time-of-use electricity price sequence, a historical elevator power sequence and a date type sequence; And inputting the multi-dimensional feature matrix into a multi-task fusion prediction model to obtain an electricity price prediction sequence and an elevator power prediction sequence in a prediction period output by the multi-task fusion prediction model.
3. The method for controlling charge and discharge of an elevator energy storage device according to claim 2, wherein the multi-task fusion prediction model comprises an input layer, a shared feature extraction layer and a task specific output head; The input layer receives the multi-dimensional feature matrix; the shared characteristic extraction layer comprises a first layer LSTM and a second layer LSTM, learns the influence of a date type on electricity price and elevator power, and outputs a hidden state sequence of each time step in the prediction period; the task specific output head comprises an electricity price prediction head and an elevator power prediction head, wherein the electricity price prediction head and the elevator power prediction head are both composed of two layers of fully-connected networks; the electricity price predicting head learns the electricity price change trend and outputs the electricity price predicting sequence; And the elevator power prediction head learns the elevator power change trend and outputs the elevator power prediction sequence.
4. The method according to claim 1, wherein the objective of maximizing the total net gain is to solve the objective function by a two-stage heuristic algorithm under the constraint of the constraint condition to obtain an optimal action sequence, and the method comprises: Generating a revenue-oriented initial motion sequence from the electricity price prediction sequence and the elevator power prediction sequence in a first stage; And in the second stage, the total net benefit is maximized, the initial action sequence is iteratively updated under the constraint of the constraint condition until the optimal action sequence is obtained by iteration convergence, wherein the constraint condition comprises the SOC range of the energy storage device, and the maximum allowable charge power and the maximum allowable discharge power under different SOCs.
5. The method of controlling charge and discharge of an elevator energy storage device of claim 4, wherein the second stage comprises: the initial action sequence is used as a current action sequence, and neighborhood operation is carried out on the current action sequence, wherein the action of one time step in the current action sequence is randomly changed, or the actions of two time steps in the current action sequence are randomly exchanged, so that a new action sequence is generated; aiming at the new action sequence, a system energy flow simulation model is called, the power grid interaction power change, the energy storage device SOC change and the energy storage device state switching event corresponding to given actions and elevator power predicted values are simulated step by step, and the total net benefit corresponding to the new action sequence is calculated; Determining whether to take the new action sequence as the optimal action sequence according to the simulated annealing probability; And if the new action sequence is not used as the optimal action sequence, the new action sequence is used as the current action sequence, and the step of executing the neighborhood operation on the current action sequence is returned to be executed until the optimal action sequence is obtained.
6. The elevator energy storage device charge-discharge control method according to any one of claims 1-5, wherein the controlling the charge-discharge of the energy storage device in the prediction period according to the optimal action sequence comprises: the first m actions in the optimal action sequence are issued to a field controller, so that the field controller executes the first m actions to control the charge and discharge of the energy storage device, and m is more than or equal to 1; And after the previous m-1 actions are completed or when the difference value between the current SOC and the predicted SOC of the energy storage device exceeds a threshold value, taking the next time step as a new prediction period starting point, and returning to execute the step of acquiring the electricity price prediction sequence and the elevator power prediction sequence in the prediction period.
7. An elevator energy storage device charge-discharge control system, comprising: The data acquisition and prediction unit is used for acquiring an electricity price prediction sequence and an elevator power prediction sequence in a prediction period; The system comprises an optimization solving unit, an action sequence optimizing model, a power price predicting sequence and an elevator power predicting sequence, wherein the action sequence optimizing model comprises an action sequence, an objective function and constraint conditions, the objective function represents total net benefits of a system after actions in the action sequence are sequentially executed on an elevator energy storage device in the predicting period, and the total net benefits are jointly determined by electricity charge benefits, aging cost of the energy storage device and abrasion cost of an energy storage device control component; and the control unit is used for controlling the charge and discharge of the energy storage device in the prediction period according to the optimal action sequence.
8. The elevator energy storage device charge and discharge control system according to claim 7, wherein the data acquisition and prediction unit is specifically configured to obtain a multidimensional feature matrix corresponding to a power grid time-of-use electricity price sequence, a historical elevator power sequence and a date type sequence, and input the multidimensional feature matrix into a multi-task fusion prediction model to obtain an electricity price prediction sequence and an elevator power prediction sequence in a prediction period output by the multi-task fusion prediction model.
9. The elevator energy storage device charge-discharge control system of claim 7, wherein the optimization solving unit comprises: a first solving subunit, configured to generate, in a first stage, a revenue-guiding initial motion sequence according to the electricity price prediction sequence and the elevator power prediction sequence; And the second solving subunit is used for iteratively updating the initial action sequence until the optimal action sequence is obtained by iteration convergence under the constraint of the constraint condition with the aim of maximizing the total net benefit in the second stage, wherein the constraint condition comprises the SOC range of the energy storage device, and the maximum allowable charging power and the maximum allowable discharging power under different SOCs.
10. The elevator energy storage device charge and discharge control system according to claim 9, wherein the second solving subunit is specifically configured to take the initial motion sequence as a current motion sequence, execute a neighborhood operation on the current motion sequence, randomly change a motion of one time step in the current motion sequence, or randomly exchange motions of two time steps in the current motion sequence, generate a new motion sequence, call a system energy flow simulation model for the new motion sequence, simulate a power grid interaction power change, an energy storage device SOC change and an energy storage device state switching event corresponding to a given motion and an elevator power predicted value step by step, calculate a total net gain corresponding to the new motion sequence, determine whether to take the new motion sequence as the optimal motion sequence according to a simulated annealing probability, and if not take the new motion sequence as the optimal motion sequence, take the new motion sequence as the current motion sequence, and return to execute a step of executing the neighborhood operation on the current motion sequence until the optimal motion sequence is obtained.
11. The charge and discharge control system of an elevator energy storage device according to any one of claims 7 to 10, wherein the optimization solving unit is further configured to issue the first m actions in the optimal action sequence to a site controller, so that the site controller executes the first m actions to perform charge and discharge control on the energy storage device, and after the first m-1 actions are completed or when a difference between a current SOC and a predicted SOC of the energy storage device exceeds a threshold value, the next time step is taken as a new prediction period start point, and trigger to execute the data acquisition and prediction unit, where m is greater than or equal to 1.

Description

Charging and discharging control method and related device for elevator energy storage device Technical Field The application relates to the technical field of elevator energy-saving control, in particular to a charge and discharge control method of an elevator energy storage device and a related device. Background The elevator can generate regenerated electric energy in the braking process, and if the regenerated electric energy is not recovered, the part of the electric energy is generally dissipated in the form of heat energy through a braking resistor, so that energy waste is caused. To improve energy efficiency, the addition of energy storage devices (e.g., batteries or supercapacitors) to the elevator system to recover and reuse this energy has become an important development direction in the industry. At present, the charging and discharging strategies of the elevator energy storage device focus on energy recovery efficiency or simple peak valley arbitrage, which may cause the elevator energy storage device to frequently switch between charging-discharging-standby states, however, the state switching of the energy storage device depends on power switching components such as contactors, relays and the like, and the frequent switching can aggravate mechanical and electrical wear of the components, thereby increasing fault risks and maintenance costs. Disclosure of Invention In view of the above problems, the application provides a charge and discharge control method of an elevator energy storage device and a related device, so as to realize the purposes of reducing the ageing of the energy storage device and the abrasion of a control component of the energy storage device into cost, optimizing the electricity charge income uniformly and improving the charge and discharge control effect of the elevator energy storage device. The specific scheme is as follows: the first aspect of the application provides a charge and discharge control method of an elevator energy storage device, comprising the following steps: acquiring an electricity price prediction sequence and an elevator power prediction sequence in a prediction period; Constructing an action sequence optimization model according to the electricity price prediction sequence and the elevator power prediction sequence, wherein the action sequence optimization model comprises an action sequence, an objective function and constraint conditions, the objective function represents total net benefit of a system after the elevator energy storage device is sequentially executed with actions in the action sequence in the prediction period, and the total net benefit is determined by electricity charge benefit, energy storage device aging cost and energy storage device control component wear cost; with the total net benefit maximization as a target, solving the objective function by adopting a two-stage heuristic algorithm under the constraint of the constraint condition to obtain an optimal action sequence; And controlling the charge and discharge of the energy storage device in the prediction period according to the optimal action sequence. In one possible implementation, the acquiring the electricity price prediction sequence and the elevator power prediction sequence in the prediction period includes: acquiring a multidimensional feature matrix corresponding to a power grid time-of-use electricity price sequence, a historical elevator power sequence and a date type sequence; And inputting the multi-dimensional feature matrix into a multi-task fusion prediction model to obtain an electricity price prediction sequence and an elevator power prediction sequence in a prediction period output by the multi-task fusion prediction model. In one possible implementation, the multi-tasking fusion prediction model includes an input layer, a shared feature extraction layer, and a task-specific output head; The input layer receives the multi-dimensional feature matrix; the shared characteristic extraction layer comprises a first layer LSTM and a second layer LSTM, learns the influence of a date type on electricity price and elevator power, and outputs a hidden state sequence of each time step in the prediction period; the task specific output head comprises an electricity price prediction head and an elevator power prediction head, wherein the electricity price prediction head and the elevator power prediction head are both composed of two layers of fully-connected networks; the electricity price predicting head learns the electricity price change trend and outputs the electricity price predicting sequence; And the elevator power prediction head learns the elevator power change trend and outputs the elevator power prediction sequence. In one possible implementation, the objective function solving by using a two-stage heuristic algorithm under the constraint of the constraint condition to obtain an optimal action sequence with the aim of maximizing the total net benefit inc