Elevator energy saving control method and elevator system

Abstract

This application discloses an elevator energy-saving control method and elevator system, relating to the field of energy storage technology. The elevator energy-saving control method is applied to an elevator energy storage system and includes: acquiring the operating data of the elevator energy storage system; dividing the operating data of each time unit into multiple time period operating data based on a preset time unit; inputting the multiple time period operating data within each time unit into a maximization objective function, and calculating a comprehensive benefit target value through the maximization objective function and its constraints; and, given target operating environment parameters, calculating corresponding control parameters based on the target operating environment parameters and the comprehensive benefit target value. This method addresses the technical problems of difficulty in adapting energy regulation control of elevator energy storage systems to different scenarios and low energy-saving efficiency, achieving adaptive matching for elevator usage scenarios of different building types, and ultimately improving the economic benefits and overall energy efficiency of the elevator energy storage system.

Description

Technical Field

[0001] This application relates to the field of energy storage technology, and in particular to an elevator energy-saving control method and elevator system. Background Technology

[0002] As the most widely distributed and frequently operated typical power equipment in commercial and public buildings, elevators account for a significant proportion of the building's total energy consumption, making them a key target for energy-saving renovations. During operation, elevator systems consume energy during heavy-load upward or acceleration phases, and regenerate energy during heavy-load downward or potential energy release phases.

[0003] Related technologies employ static, fixed battery control strategies to recover and utilize this regenerated energy, thereby achieving energy regulation and control of the elevator energy storage system. However, due to the rigid control strategies, it is difficult to adapt to diverse usage scenarios, resulting in low actual energy-saving efficiency of the energy storage system in specific scenarios. Summary of the Invention

[0004] The main purpose of this application is to provide an elevator energy-saving control method and elevator system, which aims to solve the technical problems of the difficulty in adapting the energy regulation and control of elevator energy storage systems to different scenarios and the low energy-saving efficiency.

[0005] On the one hand, an elevator energy-saving control method is provided, applied to an elevator energy storage system, the elevator energy-saving control method comprising: Obtain operational data from the elevator energy storage system; The running data for each time unit is divided into multiple time periods, using a preset time unit as the unit. The operational data of multiple time periods within each time unit are input into the objective function to maximize the overall benefit target value, and the overall benefit target value is calculated by maximizing the objective function and its constraints. Given the target operating environment parameters, the corresponding control parameters are calculated based on the target operating environment parameters and the comprehensive benefit target value.

[0006] In one embodiment, the operating data includes at least one of elevator operating mode data, operating power consumption data, battery status data, grid electricity price data, time regularity data, and control parameters to be optimized; The elevator operation mode data includes at least one of the following: average floor change, total number of elevator runs, number of elevator descents and light-load ascents indicating that the elevator energy storage system is in recyclable energy mode, and number of elevator ascents and light-load descents indicating that the elevator energy storage system is in power consumption mode; the grid electricity price data includes the current real-time electricity price and the time period identifier of the current electricity price; the time regularity data includes periodic characteristics and special day characteristics; and the control parameters to be optimized include the discharge power setpoint and the charging power threshold.

[0007] In one embodiment, dividing the running data of each time unit into multiple time period running data based on a preset time unit includes: Using a preset time unit as the basis, the operational data for each time unit is divided into K time periods. The operational data corresponding to each time period is then combined to form a feature vector, constructing a K-dimensional daily environmental feature sequence. .

[0008] In one embodiment, the step of inputting the operational data of multiple time periods within each time unit into a maximization objective function, and calculating the comprehensive benefit target value through the maximization objective function and its constraints, includes: The runtime data for multiple time periods within each time unit are input into the objective function Y=.

[0009] Define by maximizing the objective function and its constraints. Set the learning rate. , Calculate and obtain the comprehensive benefit target value under various operating environment parameters. ; The constraints are as follows: ; ; ; Define when the benefit is maximized Use the most recent Average value of environmental characteristic data for the day , ; The minimum discharge power This represents the maximum discharge power. The minimum value of the charging capacity threshold. The maximum value of the charging capacity threshold; apply a rate of change limit, and define the rate of change. , .

[0010] In one embodiment, the elevator energy-saving control method further includes the following steps: The control parameters used per unit time This is expanded to form a sequence of environmental characteristics for each unit of time. K-dimensional feature vectors with the same sequence length; Using standardized equations The continuous numerical features in the feature vector are standardized, whereby... These are the original eigenvalues. The mean of the feature in the dataset. Standard deviation; Define the objective function ,in, As a balance factor, The cost of electricity drawn from the power grid per unit time. , For power consumption, This refers to the real-time electricity price for multiple time periods within a unit of time. This represents the net change in the final state of charge of the elevator energy storage system's batteries at each unit of time. ; Define the objective function for each unit of time. ; Building a dataset ,in, Environmental characteristic sequences for each unit of time, This is a vector of control parameters for each unit of time. The comprehensive benefit target value for each unit of time; Environmental characteristic sequence at each unit time and the control parameters used per unit time Numerical values ​​are used as combined inputs to the benefit prediction model to calculate the comprehensive operational benefits per unit time. As the output of the benefit prediction model, construct the benefit prediction model. To be based on pre-trained from arrive The mapping function is used to obtain the benefit prediction model for determining the objective function. The benefit prediction model is used to determine the objective function that maximizes the benefit.

[0011] In one embodiment, the elevator energy-saving control method further includes: Update the constructed dataset to obtain a new constructed dataset: .

[0012] In one embodiment, the elevator energy-saving control method further includes: Based on the constructed dataset Benefit prediction model Train the model and define the loss function for the benefit prediction model. for ,in, , The mean squared error loss for overall revenue forecasting. The predicted value is from the benefit prediction model. As a regularization term, it is applied to the parameters of the benefit prediction model. Apply constraints, define and The weighting coefficients for the two terms are used to minimize the loss function. Update the benefit prediction model until it converges.

[0013] In one embodiment, after performing the calculation to obtain the corresponding control parameters, the process includes: The control parameters are encapsulated into structured control command frames, which are then sent to the controller of the main elevator energy storage system to control the operation of the main elevator energy storage system.

[0014] On the other hand, an elevator energy-saving control method is provided, applied to the energy storage system of the elevator body, the elevator energy-saving control method comprising: The system acquires operational data and uploads it to the elevator cloud management system. The elevator cloud management system then divides the operational data of each time unit into multiple time periods based on a preset time interval. The operational data of each time period within each time unit are input into a maximization objective function. The comprehensive benefit target value is calculated through the maximization objective function and its constraints. Based on the given target operating environment parameters, the corresponding control parameters are calculated according to the target operating environment parameters and the comprehensive benefit target value. It operates based on the control parameters issued by the elevator cloud management system.

[0015] In one embodiment, the operation of the control parameters issued by the elevator cloud management system includes: The system obtains control parameters from the elevator cloud management system, performs integrity verification and parsing on these parameters, and extracts the comprehensive benefit target value. ; In the comprehensive benefit target value Discharge power in and charging capacity threshold Within the preset operating range of the elevator energy storage system, based on the aforementioned comprehensive benefit target value... Control work.

[0016] On the other hand, an elevator system is provided, the elevator system comprising: The elevator energy storage system includes a control and management component, an elevator operation component, and a data acquisition component. The elevator operation component is used to drive the elevator to operate, and the data acquisition component is used to collect the operating data of the elevator energy storage system. An elevator cloud management system is communicatively connected to the control management component. The elevator cloud management system is configured to implement the steps of the elevator energy-saving control method described above, so as to control the operation of the elevator energy storage system through the control management component.

[0017] One or more technical solutions proposed in this application have at least the following technical effects: This method acquires operational data from an elevator energy storage system, enabling control parameter calculations to be based on actual operating scenarios while maximizing efficiency. The goal is to achieve the lowest possible grid electricity cost and maximum benefit without affecting normal elevator operation. The method divides the operational data into multiple time periods, using a preset time unit. The operational data from each time period is input into a maximization objective function, and a comprehensive benefit target value is calculated using the maximization objective function and its constraints. Based on given target operating environment parameters and the comprehensive benefit target value, corresponding control parameters are calculated. This method maximizes the comprehensive benefits of the elevator energy storage system, effectively addressing the technical problems of difficulty in adapting energy regulation and control to different scenarios and low energy efficiency. This allows for adaptive matching to elevator usage scenarios in different building types, ultimately improving the economic benefits and overall energy efficiency of the elevator energy storage system. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application 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.

[0020] Figure 1 A flowchart is provided for an embodiment of the elevator energy-saving control method applied to an elevator cloud management system according to this application; Figure 2 A schematic diagram provided for another embodiment of the elevator energy-saving control method applied to an elevator cloud management system according to this application; Figure 3 A partial flowchart is provided for another embodiment of the elevator energy-saving control method applied to an elevator cloud management system according to this application; Figure 4 A detailed flowchart is provided for one embodiment of step S300 of this application; Figure 5 This is a partial flowchart illustrating an embodiment of the elevator energy-saving control method applied to the main elevator energy storage system of this application; Figure 6 A detailed flowchart is provided for one embodiment of step S620 of this application; Figure 7 This is a schematic diagram of a module provided for an embodiment of the elevator system of this application.

[0021] Explanation of icon numbers: 100. Elevator energy storage system; 110. Control and management components; 120. Elevator operation components; 130. Data acquisition components; 131. Monitoring unit; 132. Power statistics unit; 133. Sensor data acquisition unit; 140. Battery pack; 150. Discharge unit; 160. Charging unit; 200. Elevator cloud management system; 210. Cloud data storage module; 220. Cloud data processing module; 230. Cloud model building and training module; 240. Cloud control parameter optimization and processing module.

[0022] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0023] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0024] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0025] As the most widely distributed and frequently operating typical power equipment in commercial and public buildings, elevator energy consumption accounts for a significant proportion of the building's total energy consumption, making it a key target for energy-saving renovations. Elevator systems exhibit unique bidirectional energy flow characteristics during operation: elevators consume energy during heavy-load upward movement or acceleration, and regenerate energy during heavy-load downward movement or potential energy release, thus becoming temporary energy producers. Related technologies, to effectively recover and utilize this regenerated energy, will equip elevators with local energy storage systems (including batteries). However, these technologies generally adopt static, fixed battery control strategies. While achieving basic energy recovery, they cannot fully tap the comprehensive potential of elevator systems in terms of energy saving and economy, presenting a fundamental limitation. Furthermore, the energy regulation and control of elevator energy storage systems using related technologies suffer from the following problems: First, the control strategies are rigid and difficult to adapt to diverse usage scenarios. Elevators in different types of buildings, such as office buildings, residential buildings, shopping malls, and hospitals, have fundamentally different operating load curves due to different application scenarios. Different usage habits further lead to different elevator regenerative power generation and electricity demand. However, the existing fixed strategies use the same set of control parameters (such as fixed charging power thresholds and discharge power), which cannot perceive and automatically adapt to the unique usage patterns of different buildings, and cannot optimize parameter combinations in a targeted manner, resulting in low actual energy-saving efficiency of the energy storage system in specific scenarios. Second, the use of static control parameter settings lacks a self-optimization mechanism based on operational data feedback. The key control parameters of the elevator energy storage system (such as charging power thresholds and discharge power) remain unchanged for a long time after initial setting, and cannot learn and optimize their own performance autonomously. For example, it cannot optimize and increase the discharge power from 10kW to 12kW in the next operating cycle during non-minimum electricity price periods based on past operating data. This lack of capability not only causes the system to miss opportunities to obtain higher economic benefits through dynamic adjustments, but also limits its potential to support the power grid's peak shaving and valley filling, ultimately hindering the full realization of the economic and energy-saving benefits of elevator energy storage systems. Therefore, there is an urgent need for an intelligent adaptive elevator energy-saving control method that can continuously learn and optimize control parameters to solve the technical problems of energy regulation and control of elevator energy storage systems being difficult to adapt to different scenarios and having low energy-saving efficiency.

[0026] To address the technical problems of difficulty in adapting elevator energy storage systems to different scenarios and low energy efficiency, embodiments of this application propose an elevator energy-saving control method and an elevator system. For ease of description, this application primarily focuses on control devices such as controllers, control units, and control modules with control functions as the main implementers. When the elevator energy-saving control method is applied to elevator systems and elevator energy storage systems, examples of control devices and modules such as cloud management systems and control management components of elevator energy storage systems are used for illustration.

[0027] like Figure 1 , Figure 2As shown, the elevator energy-saving control method is applied to the elevator energy storage system. The elevator energy-saving control method includes the following steps S100, S200, S300 and S400.

[0028] Step S100: Obtain the operating data of the elevator energy storage system.

[0029] For example, step S100 is used to acquire the operating data uploaded by the elevator energy storage system. The elevator energy storage system, as the main body of the elevator energy storage system, includes a data acquisition component. The data acquisition component includes a monitoring unit connected to the battery pack, a power statistics unit connected to the elevator operation component, and a sensor data acquisition unit connected to the elevator operation component, etc. The operating data of the elevator energy storage system is acquired through the data acquisition component. Among them, acquiring the operating data of the elevator energy storage system includes: acquiring at least one of the following: elevator operation mode data, operating power usage data, battery status data, grid electricity price data, time regularity data, and control parameters to be optimized.

[0030] To optimize the control strategy of the elevator energy storage system, fixed time periods can be obtained. The aforementioned various operational data (such as 1 hour or other set time periods). Among them, elevator operation mode data can be provided by sensor data acquisition unit, which is used to quantify elevator operation mode (such as recyclable energy mode, power consumption mode, etc.), load intensity and energy flow direction. Elevator operation mode data includes at least one of the following: average floor change, total number of elevator runs, number of elevator descents and light-load ascents indicating the elevator energy storage system is in recyclable energy mode, and number of elevator ascents and light-load descents indicating the elevator energy storage system is in power consumption mode. The average floor change measures the overall activity level and potential energy change scale of elevator vertical transportation. The total number of elevator runs reflects the elevator start-stop frequency and basic usage intensity. The number of elevator descents and light-load ascents indicating the elevator energy storage system is in recyclable energy mode determines the number of times the elevator motor is in energy-generating or other recyclable energy modes, which is the core energy source of the elevator energy storage system and is directly related to the estimation of recyclable energy. The number of elevator ascents and light-load descents indicating the elevator energy storage system is in power consumption mode determines the number of times the elevator motor is in power-consuming or other power-consuming modes, reflecting the main power demand of the elevator.

[0031] Battery status data is used to monitor the energy state and dynamic changes of energy storage devices such as battery packs in real time. Its core indicator can be a time period. Change in battery state of charge Battery status data indicates the change in the battery's remaining available power within a given time period. Operating power usage data, obtained from the power statistics unit, reflects the power usage of the elevator energy storage system; specific indicators are provided within a time period. The elevator energy storage system consumes electricity. .

[0032] Grid electricity price data is used to reflect external electricity price fluctuations, including current real-time electricity prices. The current electricity price is located in a specific time period (i.e., the peak, valley, or normal period). This identifier is directly associated with the established electricity price strategy and can be used to define different time period electricity price operation modes.

[0033] Time-regularity data is used to characterize the time features of data, including periodic features and special day features. Periodic features indicate the day of the week (e.g., Monday to Sunday) corresponding to the data; special day features indicate whether the data belongs to holidays.

[0034] The control parameters to be optimized include the discharge power setpoint. and charging capacity threshold In order to directly calculate the control parameters that maximize benefits based on given target operating environment parameters and comprehensive benefit target values, elevator energy storage systems can adopt different... Data collection is carried out by combining parameters, and the specific strategy is to collect data every [period]. The elevator energy storage system is operated by changing a set of control parameters every 7 days (or other set intervals). The overall data collection cycle is as follows: Days (e.g., 90 days or other set time); specifically, at least some of the aforementioned operating data or other operating data can be selected based on actual conditions, without being limited here.

[0035] Step S200: Divide the running data of each time unit into multiple time period running data, using a preset time as the unit.

[0036] After acquiring operational data, the system divides the data into multiple time periods based on preset timeframes such as days, weeks, half-weeks, half-months, and months. This allows for a more accurate matching of peak, off-peak, and valley time-of-use electricity price characteristics across multiple time periods within each time unit. Simultaneously, it fully extracts the periodic patterns of the elevator energy storage system's operation across different time units. Dividing the operational data into multiple time periods effectively reduces the interference of random fluctuations on the calculated control parameters, improving the stability and accuracy of benefit prediction. Furthermore, this setup facilitates online rolling optimization based on preset timeframes when constructing or updating the benefit prediction model used to determine the maximizing objective function, enabling the control strategy to adapt to changes in operating conditions across different time periods.

[0037] Step S300: Input the multiple time period operation data within each time unit into the maximization objective function, and calculate the comprehensive benefit target value through the maximization objective function and its constraints; Step S400: Given the target operating environment parameters, calculate the corresponding control parameters based on the target operating environment parameters and the comprehensive benefit target value.

[0038] For example, multiple time period operation data within each time unit can be directly input into the maximization objective function, and the comprehensive benefit target value can be calculated by maximizing the objective function and its constraints; alternatively, a pre-learned benefit prediction model can be called, which is used to determine the maximization objective function. Multiple time period operation data within each time unit can be input into the benefit prediction model, and the corresponding control parameters can be calculated based on the given target operating environment parameters.

[0039] Based on given target operating environment parameters, and according to the target operating environment parameters and comprehensive benefit target value, corresponding control parameters with the goal of maximizing benefits are calculated. Based on these control parameters, the operation of the elevator energy storage system is further controlled. This is used to dynamically solve for the comprehensive benefit target value (optimal control parameters) based on the acquired elevator operation mode data, operating power usage data, battery status data, grid electricity price data, time regularity data, control parameters to be optimized, and other operating data, as well as the given target operating environment parameters. This enables adaptive adjustment of the control strategy, thereby reasonably setting control parameters such as charging power threshold and discharge power when the elevator regenerative energy is sufficient, and maximizing the recovery of regenerative braking energy for charging during off-peak hours and discharging during peak hours, while taking into account battery safety and lifespan, and reducing losses caused by overcharging, over-discharging, and frequent high-power charging and discharging.

[0040] In the embodiments of this application, step S100 comprehensively acquires the operating data of the elevator energy storage system to determine the multi-dimensional state of the elevator energy storage system, thereby clarifying the recyclable energy mode and power consumption mode of the elevator energy storage system. This allows the calculation of control parameters to be based on the operating data of the actual operating scenario, while taking into account the maximization of benefits. This is used to minimize the grid electricity price cost and maximize the benefits without affecting the normal operation of the elevator. Steps S200, S300, and S400 calculate control parameters to maximize benefits and control the elevator energy storage system based on the optimized parameters obtained from the calculations. This aims to maximize the overall benefits of the elevator energy storage system. The maximization objective function can cover multiple aspects, including regenerative energy recovery, maximizing benefits based on peak-valley electricity prices, battery life protection, and grid load optimization. For example, based on elevator operation mode data, future electricity demand and the elevator's own power generation potential are predicted. When the elevator generates more electricity during descent, regenerative electricity is prioritized to charge the battery. Based on operating electricity usage data, the battery is discharged in advance to supply power to the elevator during peak grid electricity prices or peak elevator electricity consumption. Charging and discharging are controlled based on battery state data to stop charging when the SOC is too high and limit discharging when the SOC is too low. Electricity prices are predicted based on grid electricity price data to prioritize discharging when prices are high and prioritizing charging when prices are low. In this way, the energy consumption and electricity costs of elevator operation can be effectively reduced, while improving the stability and reliability of system operation.

[0041] This application addresses the shortcomings of static and fixed battery control strategies used in elevator energy storage systems in related technologies by proposing a parameter adaptive optimization elevator energy-saving control method and elevator system. By continuously collecting operational data from the elevator energy storage system, the system acquires self-sensing, self-learning, and self-optimization capabilities, effectively solving the technical problems of difficulty in adapting energy regulation and control of elevator energy storage systems to different scenarios and low energy-saving efficiency. Based on historical operational data, the elevator system can dynamically adjust key control parameters such as charging power threshold and discharging power in the battery control strategy for the next cycle, thereby achieving adaptive matching for elevator usage scenarios of different building types, ultimately improving the economic benefits and overall energy efficiency of the elevator energy storage system.

[0042] In one embodiment, step S200, dividing the running data of each time unit into multiple time period running data based on a preset time unit, includes: Using a preset time unit as the basis, the operational data for each time unit is divided into K time periods. The operational data corresponding to each time period is then combined to form a feature vector, constructing a K-dimensional daily environmental feature sequence. .

[0043] For example, embodiments of this application can convert acquired historical operational data into a standardized, structured dataset, which can be stored in the cloud storage device of the elevator cloud management system or other cloud data storage modules. For instance, with a preset time period of days, using days as the basic processing unit, the operational data for each time unit is divided into K time-segment operational data. (For example, if a time period is 1 hour, then) ).

[0044] like Figure 3 As shown, in one embodiment, the elevator energy-saving control method further includes the following steps: A benefit prediction model is constructed, which is used to determine the objective function to maximize the benefit.

[0045] Specifically, it includes the following steps: Step S510: The control parameters used for each unit of time... This is expanded to form a sequence of environmental characteristics for each unit of time. The sequence consists of K-dimensional feature vectors of the same length. By constructing time-series feature vectors and aligning control parameters, the consistency and integrity of the time-dimensional data are ensured to a certain extent.

[0046] Step S520: Use standardized equations The continuous numerical features in the feature vector are standardized, whereby... These are the original eigenvalues. The mean of the feature in the dataset. Standard deviation Standardization methods can eliminate dimensional differences in continuous numerical features, improve data compatibility, and thus achieve adaptive matching for elevator usage scenarios in different building types, ultimately improving the economic benefits and overall energy efficiency of elevator energy storage systems.

[0047] Step S530: Define the objective function ,in, As a balance factor, The cost of electricity drawn from the power grid per unit time. , For power consumption, This refers to the real-time electricity price for multiple time periods within a unit of time. This represents the net change in the final state of charge of the elevator energy storage system's batteries at each unit of time.

[0048] Step S540: Define the maximization objective function for each unit of time. The benefit target value is calculated based on maximizing the objective function, and the benefit target value adopts... The maximum value is obtained, which is equivalent to minimizing the cost of electricity used by the grid, and ultimately resulting in more surplus electricity each day.

[0049] Step S550: Construct the dataset ,in, Environmental characteristic sequences for each unit of time, This is a vector of control parameters for each unit of time. The comprehensive benefit target value for each unit of time; Step S560: Using the environmental feature sequence for each unit time. and the control parameters used per unit time Numerical values ​​are used as combined inputs to the benefit prediction model to calculate the comprehensive operational benefits per unit time. As the output of the benefit prediction model, construct the benefit prediction model. To be based on pre-trained from arrive The mapping function is used to obtain the benefit prediction model for determining the objective function. The benefit prediction model is used to determine the objective function that maximizes the benefit.

[0050] Understandably, embodiments of this application convert historical operating data collected by the elevator energy storage system into a standardized, structured dataset suitable for training a benefit prediction model. This dataset can be stored in a cloud storage device or other cloud data storage module. Taking a preset time period of one day as an example, the day is divided into consecutive... This divides the operational data for each time unit into K time periods. (For example, if a time period is 1 hour, then) For each time period, its corresponding elevator operation mode data, power grid price data, and time pattern data are combined into a feature vector. This process is then applied to all elevators throughout the day. A time period can be used to construct a The daily environmental feature sequence of dimension is denoted as To ensure data consistency across time, the control parameters used on that day... The values ​​are copied and expanded to form a result similar to... Same sequence length This allows for the creation of 3D vectors, thus completing the feature alignment process.

[0051] To eliminate dimensional differences, continuous numerical features in the feature vector can be standardized using methods such as Z-score standardization or other standardization techniques. Taking Z-score standardization as an example, the standardization equation is... ,in, These are the original eigenvalues. The mean of the feature in the dataset. The standard deviation is used to eliminate dimensional differences.

[0052] To quantify the overall operational benefits of the system under specific environments and control strategies, an objective function is defined. The calculation formula is as follows: ,in, The cost of electricity from the Japanese power grid, This reflects the total cost of electricity consumed by the elevator system from the power grid after applying the current control strategy. The calculation formula is: ,in, For power consumption, This refers to the real-time electricity price. This represents the net change in the final state of charge of the battery each day, reflecting the final change in the energy stored in the battery after a day of operation, and indicating the flow of some energy. The calculation formula is: ,in, This is a balancing factor used to adjust the relative weights of grid electricity costs and battery energy storage state changes in the objective function. Furthermore, the maximization objective function for each unit of time is defined. The benefit target value is calculated based on maximizing the objective function, and the benefit target value adopts... The maximum value is obtained, which is equivalent to minimizing the cost of electricity used by the grid, and ultimately resulting in more surplus electricity each day. The larger the value, the better the economic benefits the system can obtain by adopting the current control strategy under the same elevator operating environment and electricity price conditions.

[0053] The data processed as described above are grouped into standard samples by day, with each sample consisting of a triplet: .in, This is the environmental characteristic sequence for that day. This is the control parameter vector for that day. The comprehensive benefit target value calculated for that day forms an ordered dataset from the operational data. To ensure data timeliness, a rolling update mechanism can be used to automatically replace old samples. Let the data window size be... Heaven, the first The dataset for the day is .

[0054] Employing a rolling update mechanism ensures continuous updates to the dataset, maintaining data timeliness and ultimately creating a standardized cloud dataset directly usable for training benefit prediction models. For example, iterative training can converge the benefit prediction model, maximizing benefits, and control parameters can be optimized by incorporating factors such as battery safety and grid constraints. The dataset undergoes rolling updates, automatically replacing older data samples to ensure the data remains up-to-date. Let the data window size be... Heavens, then the first The dataset for the day is: When new runtime data arrives, a rolling update is performed: .

[0055] The embodiments of this application employ a rolling update mechanism with a set time window to dynamically maintain the dataset. By removing the oldest historical running data samples and incorporating the most recently acquired running data samples, while maintaining a stable dataset size, the continuous injection of new running data reflecting the current system and environmental characteristics reduces the computational burden of training and minimizes interference from outdated samples. This provides a high-quality data foundation that is timely and scalable for subsequent training of the benefit prediction model. In this way, a reliable basis can be provided for optimizing control parameters, enabling the system to continuously formulate optimal operating strategies in a dynamically changing environment and achieve long-term stable profit maximization control.

[0056] The embodiments of this application also include the following steps: Based on the constructed dataset Benefit prediction model Train the model and define the loss function for the benefit prediction model. for ,in, , The mean squared error loss for overall revenue forecasting. The predicted value is from the benefit prediction model. As a regularization term, it is applied to the parameters of the benefit prediction model. Apply constraints, define and The weighting coefficients for the two terms are used to minimize the loss function. Update the benefit prediction model until it converges.

[0057] The embodiments of this application establish a mathematical model capable of characterizing the complex mapping relationship between control parameters, operating scenarios, and comprehensive benefits as a benefit prediction model, used for subsequent control parameter calculation and optimization. For example, the benefit prediction model uses a daily environmental feature sequence obtained through preprocessing. and the control parameter vector applied on the same day As a combined input to the model, the calculated daily comprehensive benefit target value is obtained. As the output of the benefit prediction model, a neural network model is constructed using deep learning techniques. As a benefit prediction model, this benefit prediction model is used to determine the objective function to maximize, and is trained to learn from... arrive The mapping function enables it to learn different external environments. Below, control parameters using different control strategies At that time, the overall benefits ultimately presented by the system The changing pattern.

[0058] The benefit prediction model uses the dataset constructed using the aforementioned steps. Train the model and define its loss function. for .in, , The mean squared error loss for overall revenue forecasting. These are the predicted values ​​from the benefit prediction model; As a regularization term, it is applied to the model parameters. Imposing constraints prevents overfitting under limited data and improves the generalization ability of the benefit prediction model. and These are the weighting coefficients for the two terms. The loss function is minimized using optimization methods such as backpropagation. The network parameters are iteratively updated until the model converges.

[0059] For example, in the initial stage of model training, multiple elevator scenario data can be used for pre-training, and finally, the benefit prediction model can be fine-tuned using operational data from a specific elevator scenario. The training and update cycle of this benefit prediction model is the same as the system data update cycle. Keep in sync. Each time the rolling database completes a cycle... After the running data is updated, the model is tested using the latest dataset. Perform a retraining or fine-tuning update. The resulting benefit prediction model after training convergence. This is the benefit prediction model required for this application, and its mathematical relationship can be expressed as follows: .

[0060] The embodiments of this application employ a strategy of multi-scenario pre-training combined with scenario-specific fine-tuning, and integrate the update cycle of the benefit prediction model with the system data update cycle. By maintaining synchronization, a benefit prediction model that can be used to optimize the control parameters of elevator energy storage systems can be ultimately formed. To adapt to the energy regulation requirements of different scenarios, effectively improve energy efficiency, and achieve the maximum benefit operation control of the elevator system.

[0061] like Figure 4 As shown, in one embodiment, step S300, inputting the multiple time period operation data within each time unit into the maximization objective function, and calculating the comprehensive benefit target value through the maximization objective function and its constraints, includes: Step S310: Input the running data of multiple time periods within each time unit into the objective function Y= : Step S320: Define the objective function and its constraints by maximizing the objective function. Set the learning rate. , Calculate and obtain the comprehensive benefit target value under various operating environment parameters. .

[0062] The constraints are as follows: ; ; ; Define when the benefit is maximized Use the most recent Average value of environmental characteristic data for the day , ; The minimum discharge power This represents the maximum discharge power. The minimum value of the charging capacity threshold. The maximum value of the charging capacity threshold; apply a rate of change limit, and define the rate of change. , .

[0063] Understandably, the operational data from multiple time periods within each time unit can be directly input into the objective function to maximize the overall benefit target value, which can then be calculated using the objective function and its constraints. Alternatively, a pre-learned benefit prediction model can be invoked. This model determines the objective function, and the operational data from multiple time periods within each time unit can be input into the model. Based on given target operating environment parameters, the corresponding control parameters can be calculated. In this way, the overall benefit target value that maximizes the predicted overall benefit of the system under future operating scenarios can be determined. For given target operating environment parameters... To find the comprehensive benefit target value The problem can be defined as the following optimization problem: the objective function is Y= , Among them, the objective function Represents the expected environment The following control parameters are used The model predicts the overall benefit. Constraints include the physical feasible region boundaries of the control parameters. and other constraints .

[0064] Due to the specific environmental characteristics of the future Since the decision-making time is unknown, in some alternative implementations, the historical mean forecasting method is used, i.e., using the most recent... Day (for example) The average value of environmental characteristic data is used as The estimated value: .

[0065] This method is based on the reasonable assumption that the operating environment parameters have short-term stationarity or periodicity, and uses historical average patterns to approximate typical future scenarios, providing a definite input for the optimization problem. Alternatively, environmental characteristics of dates in historical operating data that have the same attributes as the target date (e.g., all are Mondays) can be selected as estimates; this is not limited here.

[0066] To ensure that control parameters remain within the safe and feasible range allowed by the system, boundary constraints are imposed, among which... and These are the minimum and maximum values ​​of the discharge power, respectively. and These represent the minimum and maximum values ​​of the charging capacity threshold, respectively. Smoothness constraints are also applied to prevent excessive adjustments in control parameters from causing system oscillations; a rate-of-change limit is imposed. ; That is, limiting the discharge power relative to the current value The maximum adjustment range limits the charging capacity threshold relative to the current value. The maximum adjustment range.

[0067] For example, a gradient optimization method is used to solve the problem. The differentiability of the neural network is utilized to calculate the objective function. For control parameters gradient: , The gradient ascent method is used to search for and solve the problem. , ,in This is the learning rate.

[0068] As an alternative, Bayesian optimization and other optimization methods can be used to solve for the comprehensive benefit target value. No further restrictions are imposed here.

[0069] In one embodiment, after performing step S400 and calculating the corresponding control parameters, the process includes: The control parameters are encapsulated into structured control command frames, which are then sent to the controller of the main elevator energy storage system to control the operation of the main elevator energy storage system.

[0070] Perform parameter optimization calculations and obtain comprehensive benefit target values ​​in the cloud. Subsequently, the cloud optimization server of the cloud management system encapsulates these control parameters into structured control command frames. These encapsulated control command frames are then transmitted via reliable communication links such as wireless cellular networks or other communication components to the controller or other control management components of the target elevator's energy storage system. Taking the controller of the elevator's energy storage system as an example, the controller's built-in command receiving and parsing module continuously monitors the communication port, completing the reception, integrity verification, and parsing of the command frames, ultimately extracting the comprehensive benefit target value. In this way, the comprehensive benefit target value calculated by cloud optimization can be safely and reliably deployed to the energy storage control system of the elevator body, and these control parameters can be correctly applied in subsequent operation.

[0071] For example, before applying new control parameters, the control and management components of an elevator energy storage system will first determine the received discharge power. and charging capacity threshold Is it within the preset operating range of the elevator's energy storage system? If it is within the preset range, the control and management components will adjust the operating parameters accordingly during subsequent operation. The control strategy is then executed. Afterward, the system enters a new operational data acquisition cycle, continuously recording and uploading operational data to provide feedback for the next round of model learning and control parameter calculation. This forms a complete adaptive closed loop, from the perception of operational data to the learning and optimization of the benefit prediction model, and then to the application of the benefit prediction model and the updating and optimization of control parameters.

[0072] This application provides an intelligent adaptive elevator energy-saving control method capable of continuously learning and optimizing control parameters. Compared to the static, fixed battery control strategies widely used in related technologies, this application constructs a complete adaptive closed loop, from the perception of operating data to the learning and optimization of the benefit prediction model, and then to the application of the benefit prediction model and the updating and optimization of control parameters. This allows the system to autonomously extract knowledge from its own historical operating data, update the cognitive model during the pre-training of the benefit prediction model, achieve adaptive calculation and optimization of parameters, and execute a better control strategy accordingly. This application can automatically identify and adapt to the elevator operating modes and usage habits of vastly different application scenarios such as office buildings, residences, and shopping malls, and call up matching comprehensive benefit target value combinations, thereby improving the energy-saving effect and economic benefits of the entire elevator energy storage system, systematically reducing building energy consumption, and more effectively supporting the power grid's "peak shaving and valley filling," providing a new intelligent and adaptive technical path for achieving energy conservation.

[0073] like Figure 5 As shown, one embodiment of this application provides an elevator energy-saving control method, which is applied to the elevator's energy storage system. The elevator energy-saving control method includes: Step S610: Acquire operating data and upload the acquired operating data to the elevator cloud management system, so that the elevator cloud management system divides the operating data of each time unit into multiple time period operating data based on a preset time unit; inputs the multiple time period operating data within each time unit into the maximization objective function, calculates the comprehensive benefit target value through the maximization objective function and its constraints; and calculates the corresponding control parameters based on the given target operating environment parameters and the comprehensive benefit target value. Step S620: Work based on the control parameters issued by the elevator cloud management system.

[0074] The elevator energy storage system includes a data acquisition component, which comprises a monitoring unit connected to the battery pack, a power statistics unit connected to the elevator operation component, and a sensor data acquisition unit connected to the elevator operation component. The data acquisition component acquires the operating data of the elevator energy storage system. The elevator cloud management system can calculate the corresponding control parameters using the steps described in the preceding embodiments. The cloud optimization server of the cloud management system encapsulates these control parameters into structured control command frames. These encapsulated control command frames are then transmitted to the controller or other control management components of the target elevator's elevator energy storage system via a reliable communication link such as a wireless cellular network or other communication components.

[0075] like Figure 6 As shown, for example, step S620, the operation based on the control parameters issued by the elevator cloud management system, includes: Step S621: Obtain the control parameters issued by the elevator cloud management system, perform integrity verification and parsing on the control parameters, and extract the comprehensive benefit target value. ; Step S622: At the comprehensive benefit target value Discharge power in and charging capacity threshold Within the preset operating range of the elevator energy storage system, based on the aforementioned comprehensive benefit target value... Control work.

[0076] Taking the controller of the elevator energy storage system as an example, the built-in command receiving and parsing module continuously monitors the communication port, completes the receiving, integrity verification and parsing of command frames, and finally extracts the comprehensive benefit target value. In this way, the comprehensive benefit target value calculated in the cloud can be safely and reliably deployed to the elevator's energy storage control system, ensuring that these control parameters are correctly applied in subsequent operation. Before applying new control parameters, the elevator's energy storage system's control and management components and other onboard control and management units will first determine the received discharge power. and charging capacity threshold Is it within the preset operating range of the elevator's energy storage system? If it is within the preset range, the control and management components will adjust the operating parameters accordingly during subsequent operation. The control strategy is then executed. Afterward, the system enters a new operational data acquisition cycle, continuously recording and uploading operational data to provide feedback for the next round of model learning and control parameter calculation. This forms a complete adaptive closed loop, from the perception of operational data to the learning and optimization of the benefit prediction model, and then to the application of the benefit prediction model and the updating and optimization of control parameters.

[0077] The elevator energy-saving control method applied to the main body elevator energy storage system in this embodiment can solve the technical problems of difficulty in adapting the energy regulation and control of the elevator energy storage system to different scenarios and low energy-saving efficiency. Compared with the prior art, the beneficial effects of the elevator system applied to the main body elevator energy storage system provided in this application are the same as the beneficial effects of the elevator energy-saving control method applied to the elevator cloud management system provided in the above embodiment, and other technical features of the elevator system are the same as those disclosed in the above embodiment method, and will not be repeated here.

[0078] like Figure 7 As shown, one embodiment of this application provides an elevator system, which includes a main elevator energy storage system 100 and an elevator cloud management system 200.

[0079] The elevator energy storage system 100 includes a control management component 110, an elevator operation component 120, and a data acquisition component 130. The elevator operation component 120 is used to drive the elevator, and may include, for example, the entire system equipment that drives the elevator (such as a frequency converter, motor, etc.). The data acquisition component 130 is used to collect the operating data of the elevator energy storage system 100. The elevator cloud management system 200 is communicatively connected to the control management component 110. The elevator cloud management system 200 is configured to implement the steps of the elevator energy-saving control method as described in the above embodiment, so as to control the operation of the elevator energy storage system 100 through the control management component 110.

[0080] For example, the main body elevator energy storage system 100, as the main body elevator energy storage system 100, also includes a battery pack 140 or other energy storage device, a discharge unit 150, and a charging unit 160. The battery pack 140 is an energy storage and rechargeable device, which can be composed of multiple batteries connected in series; the output terminal of the charging unit 160 is connected to the battery pack 140, the input terminal of the discharge unit 150 is connected to the battery pack 140, and the output terminal of the discharge unit 150 is connected to the elevator operating component 120; the acquisition component 130 includes a monitoring unit 131 connected to the battery pack 140, a power statistics unit 132 connected to the elevator operating component 120, and a sensor data acquisition unit 133 connected to the elevator operating component 120. The monitoring unit 131 is used to monitor the information of each module of the battery pack 140, such as temperature, voltage, and alarm information; the charging unit 160 and the discharging unit 150 are used to control the start of battery charging and discharging and power adjustment; the power statistics unit 132 is used to count the power usage of the entire elevator energy storage system 100 (e.g., electricity meter); the sensor data acquisition unit 133 is used to collect sensor data such as elevator load, direction, and speed. The control and management component 110 is used to monitor the operating status of the battery, inverter, DC bus circuit and electrical equipment, etc. It can control the charging and discharging of the battery, acquire the power data of the power statistics unit 132 and the data collected by the elevator sensor data acquisition unit 133, thereby acquiring at least one or more of the following operating data of the main elevator energy storage system 100: elevator operation mode data, power usage data, battery status data, grid electricity price data, time regularity data, and control parameters to be optimized. It can also upload the acquired operating data to the elevator cloud management system 200, and can realize the energy-saving control and system safety protection functions of the main elevator energy storage system 100 according to the control parameters and other related control strategies optimized by the aforementioned elevator energy-saving control method.

[0081] The elevator cloud management system 200 includes a cloud data storage module 210, a cloud data processing module 220, a cloud model building and training module 230, and a cloud control parameter optimization processing module 240. The cloud data storage module 210 stores the operating data of the main elevator energy storage system 100 uploaded by the control management component 110; the cloud data processing module 220 can perform special processing on the acquired operating data according to a predetermined method and construct a dataset; the cloud model building and training module 230 is used to model the main elevator energy storage system 100 and train a benefit prediction model; the cloud parameter optimization processing module uses the trained benefit prediction model to optimize specified control parameters and sends the optimized control parameters to the control management component 110, so that the control management component 110 can control the operation of the main elevator energy storage system 100.

[0082] The elevator system provided in this application, employing the elevator energy-saving control method described in the above embodiments, can solve the technical problems of difficulty in adapting the energy regulation and control of elevator energy storage systems to different scenarios and low energy-saving efficiency. Compared with the prior art, the beneficial effects of the elevator system provided in this application are the same as those of the elevator energy-saving control method provided in the above embodiments, and other technical features of the elevator system are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0083] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. An elevator energy-saving control method, applied to an elevator cloud management system, characterized in that, The elevator energy-saving control method includes: Obtain operational data from the elevator energy storage system; The running data for each time unit is divided into multiple time periods, using a preset time unit as the unit. The operational data of multiple time periods within each time unit are input into the objective function to maximize the overall benefit target value, and the overall benefit target value is calculated by maximizing the objective function and its constraints. Given the target operating environment parameters, the corresponding control parameters are calculated based on the target operating environment parameters and the comprehensive benefit target value.

2. The elevator energy-saving control method as described in claim 1, characterized in that, The operational data includes at least one of the following: elevator operation mode data, power consumption data, battery status data, grid electricity price data, time-based regularity data, and control parameters to be optimized. The elevator operation mode data includes at least one of the following: average floor change, total number of elevator runs, number of elevator descents and light-load ascents indicating that the elevator energy storage system is in recyclable energy mode, and number of elevator ascents and light-load descents indicating that the elevator energy storage system is in power consumption mode; the grid electricity price data includes the current real-time electricity price and the time period identifier of the current electricity price; the time regularity data includes periodic characteristics and special day characteristics; and the control parameters to be optimized include the discharge power setpoint and the charging power threshold.

3. The elevator energy-saving control method as described in claim 1, characterized in that, The process of dividing the operational data of each time unit into multiple time periods, using a preset time unit, includes: Using a preset time unit as the basis, the operational data for each time unit is divided into K time periods. The operational data corresponding to each time period is then combined to form a feature vector, constructing a K-dimensional daily environmental feature sequence. .

4. The elevator energy-saving control method as described in claim 1, characterized in that, The step of inputting the operational data of multiple time periods within each time unit into the maximization objective function, and calculating the comprehensive benefit target value through the maximization objective function and its constraints, includes: The runtime data for multiple time periods within each time unit are input into the objective function Y=. Define by maximizing the objective function and its constraints. Set the learning rate. , Calculate and obtain the comprehensive benefit target value under various operating environment parameters. ; The constraints are as follows: ; ; ; Define when the benefit is maximized Use the most recent Average value of environmental characteristic data for the day , ; The minimum discharge power This represents the maximum discharge power. The minimum value of the charging capacity threshold. The maximum value of the charging capacity threshold; apply a rate of change limit, and define the rate of change. , .

5. The elevator energy-saving control method as described in claim 1, characterized in that, The elevator energy-saving control method also includes the following steps: The control parameters used per unit time This is expanded to form a sequence of environmental characteristics for each unit of time. K-dimensional feature vectors with the same sequence length; Using standardized equations The continuous numerical features in the feature vector are standardized, whereby... These are the original eigenvalues. The mean of the feature in the dataset. Standard deviation; Define the objective function ,in, As a balance factor, The cost of electricity drawn from the power grid per unit time. , For power consumption, This refers to the real-time electricity price for multiple time periods within a unit of time. This represents the net change in the final state of charge of the elevator energy storage system's batteries at each unit of time. ; Define the objective function for each unit of time. ; Building a dataset ,in, This is a sequence of environmental characteristics for each unit of time. This is a vector of control parameters for each unit of time. The comprehensive benefit target value for each unit of time; Environmental characteristic sequence at each unit time and the control parameters used per unit time Numerical values ​​are used as combined inputs to the benefit prediction model to calculate the comprehensive operational benefits per unit time. As the output of the benefit prediction model, construct the benefit prediction model. To be based on pre-trained from arrive The mapping function is used to obtain the benefit prediction model for determining the objective function. The benefit prediction model is used to determine the objective function that maximizes the benefit.

6. The elevator energy-saving control method as described in claim 5, characterized in that, The elevator energy-saving control method also includes: Update the constructed dataset to obtain a new constructed dataset: 。 7. The elevator energy-saving control method as described in claim 5, characterized in that, The elevator energy-saving control method also includes: Based on the constructed dataset Benefit prediction model Train the model and define the loss function for the benefit prediction model. for ,in, , The mean squared error loss for overall revenue forecasting. The predicted value is from the benefit prediction model. As a regularization term, it is applied to the parameters of the benefit prediction model. Apply constraints, define and The weighting coefficients for the two terms are used to minimize the loss function. Update the benefit prediction model until it converges.

8. The elevator energy-saving control method according to any one of claims 1 to 7, characterized in that, After performing the calculation to obtain the corresponding control parameters, the process includes: The control parameters are encapsulated into structured control command frames, which are then sent to the controller of the main elevator energy storage system to control the operation of the main elevator energy storage system.

9. An elevator energy-saving control method, applied to the energy storage system of the elevator body, characterized in that, The elevator energy-saving control method includes: The system acquires operational data and uploads it to the elevator cloud management system. The elevator cloud management system then divides the operational data of each time unit into multiple time periods based on a preset time interval. The operational data of each time period within each time unit are input into a maximization objective function. The comprehensive benefit target value is calculated through the maximization objective function and its constraints. Based on the given target operating environment parameters, the corresponding control parameters are calculated according to the target operating environment parameters and the comprehensive benefit target value. It operates based on the control parameters issued by the elevator cloud management system.

10. The elevator energy-saving control method as described in claim 9, characterized in that, The operation based on the control parameters issued by the elevator cloud management system includes: The system obtains control parameters from the elevator cloud management system, performs integrity verification and parsing on these parameters, and extracts the comprehensive benefit target value. ; In the comprehensive benefit target value Discharge power in and charging capacity threshold Within the preset operating range of the elevator energy storage system, based on the aforementioned comprehensive benefit target value... Control work.

11. An elevator system, characterized in that, The elevator system includes: The elevator energy storage system includes a control and management component, an elevator operation component, and a data acquisition component. The elevator operation component is used to drive the elevator to operate, and the data acquisition component is used to collect the operating data of the elevator energy storage system. An elevator cloud management system is communicatively connected to the control management component. The elevator cloud management system is configured to implement the steps of the elevator energy-saving control method as described in any one of claims 1 to 8, so as to control the operation of the elevator energy storage system through the control management component.

Images