Shared elevator energy-saving and energy-storing system and method based on energy feedback and cooperative scheduling
By using a collaborative architecture between a local edge controller and a cloud dispatch center, along with a hybrid energy storage system, the problems of wasted renewable energy in elevators and low utilization of energy storage resources have been solved. This has enabled efficient energy management and safety assurance for elevator systems, while reducing energy consumption and interaction costs.
Patent Information
- Application Number
- CN202511625488.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-07
- Publication Date
- 2026-02-17
AI Technical Summary
Existing elevator systems suffer from problems such as waste of regenerated energy, lack of coordination between energy storage and dispatching, low utilization rate of energy storage resources, and insufficient safety, resulting in high energy consumption, high interaction costs, and prominent safety risks.
The system adopts a two-tier architecture of local edge controller and cloud dispatch center, combined with big data analysis and machine learning, to achieve full recovery and efficient utilization of elevator regenerative energy. Through a common DC bus and hybrid energy storage system, it enables coordinated scheduling and shared utilization of energy, and builds a multi-layered security defense to ensure system stability.
It achieves efficient recovery and utilization of elevator regenerated energy, reduces grid interaction costs, improves the utilization rate and safety of energy storage systems, extends the lifespan of elevator electrical components, and enhances system adaptability and safety assurance.
Smart Images
Figure CN121546665A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of elevator technology, energy storage technology, power systems and their automation, cloud computing and artificial intelligence, specifically to a shared elevator energy-saving and energy storage system and method based on energy feedback and collaborative scheduling. Background Technology
[0002] With the acceleration of urbanization, elevators have become a core transportation device in high-rise buildings. However, their operation suffers from significant energy waste and resource imbalance, and current technologies struggle to overcome the following bottlenecks: Regenerative energy is wasted: When elevators brake or descend under heavy load, a large amount of regenerative energy is generated (accounting for 20%-40% of the total energy consumption of elevators). Traditional systems convert this into heat energy dissipation through braking resistors, which not only wastes energy but also causes the machine room temperature to rise, resulting in additional air conditioning heat dissipation energy consumption. A few elevators equipped with independent energy storage units only serve a single elevator due to the lack of coordinated scheduling, resulting in an idle rate of over 60%.
[0003] Lack of coordination between energy storage and dispatch: There is no data interaction and coordinated control between elevator energy storage and the power grid or other elevators. The charging and discharging of energy storage depends entirely on the status of a single elevator and cannot be optimized by taking into account external conditions such as electricity price fluctuations (peak-valley price difference can be 3-5 times) and peak-valley load of the power grid. For example, the energy storage is fully charged during the peak period but cannot feed back energy to the power grid. During the off-peak period, the elevator regenerates insufficient energy but still needs to purchase electricity from the power grid at a high price, resulting in high power grid interaction costs.
[0004] Low utilization rate and limited functionality of energy storage resources: Most existing elevator energy storage systems are "dedicated" and are only used to smooth out energy fluctuations in a single elevator, without realizing energy storage resource sharing; in office buildings and other scenarios, elevator usage is low during non-office hours, and energy storage resources are idle for a long time, unable to provide power to surrounding businesses, emergency loads and other external users, thus failing to realize the added value potential of energy storage.
[0005] Insufficient control precision and safety assurance: Traditional dispatching adopts a "fixed value triggering" mode, which lacks prediction-based dynamic control. The voltage fluctuation of the common DC bus can reach ±10% of the rated value, affecting the stability of elevator operation. The energy storage system only relies on simple overvoltage protection and does not combine multiple dimensions such as temperature, SOC (state of charge), and depth of charge and discharge to build a safety defense. The lifespan degradation problem caused by overcharging and over-discharging of high-nickel batteries or solid-state batteries is prominent (cycle life is shortened by more than 40%).
[0006] The aforementioned problems result in high energy consumption during elevator operation, waste of energy storage resources, high grid interaction costs, and prominent safety risks, making it impossible to meet energy-saving requirements and energy sharing needs. Therefore, there is an urgent need for an integrated system and method that combines energy recovery, coordinated scheduling, shared utilization, and safety protection. Summary of the Invention
[0007] The purpose of this invention is to provide a shared elevator energy-saving and energy storage system and method based on energy feedback and collaborative scheduling to solve the problems mentioned in the background art.
[0008] To achieve the above objectives, the present invention provides the following technical solution: a shared elevator energy-saving and energy storage system based on energy feedback and collaborative scheduling, comprising a local edge controller (ECC), a cloud scheduling center, a field device layer, and a shared energy storage system; The local edge controller (ECC) is used to collect elevator operating status parameters, energy storage system state of charge (SOC) and grid electricity price signals, upload real-time data to the cloud dispatch center, and receive optimization strategies issued by the cloud dispatch center to perform real-time energy management and scheduling, as well as strategy execution and safety protection. The cloud dispatch center is equipped with a big data analysis and machine learning module and a multi-objective optimization dispatch module. The big data analysis and machine learning module learns from historical data on electricity price trends, load patterns and elevator usage patterns. The multi-objective optimization dispatch module aims to achieve the best energy efficiency of the elevator system, the lowest grid interaction cost and the highest utilization rate of shared energy storage. It constructs a multi-objective optimization function and solves it to generate an optimization strategy. The field equipment layer includes an elevator group, frequency converters corresponding to each elevator, and a common DC bus. The frequency converter of each elevator is connected to the common DC bus through a DC / DC module, which is used to transmit the regenerative power generated when the elevator brakes or is under heavy load during descent to the common DC bus, or to obtain power from the common DC bus to drive the elevator to run. The shared energy storage system includes a bidirectional DC / DC converter, modular energy storage units, a battery management system (BMS), and a bidirectional AC / AC converter. The bidirectional DC / DC converter is connected to a common DC bus and is used to regulate the voltage of the common DC bus and realize bidirectional energy flow between the common DC bus and the modular energy storage units. The modular energy storage units are composed of at least one energy storage medium selected from lithium batteries, lead-acid batteries or high-nickel batteries, solid-state batteries, and supercapacitors. The battery management system (BMS) is used to monitor the voltage, current, temperature, and SOC of each energy storage medium and to perform charge and discharge protection. The bidirectional AC / AC converter connects the modular energy storage units to the national power grid and is used to realize bidirectional AC power interaction between the energy storage system and the power grid, while supporting the grid to charge the energy storage system and the energy storage system to feed back power to the grid.
[0009] Preferably, the multi-objective optimization function constructed by the multi-objective optimization scheduling module of the cloud scheduling center is:
[0010] in The combined energy efficiency function of the elevator system and the energy storage system is expressed as follows:
[0011] in This refers to the amount of regenerated electrical energy recovered by the elevator. The electricity provided by shared energy storage to the outside world Electrical energy input from the power grid, This refers to system energy loss. in The power grid interaction cost function is expressed as follows:
[0012] in for The power exchange with the power grid at any time (positive input, negative output). for The grid electricity price at any given time, where T is the dispatch period; in The shared energy storage utilization rate function is expressed as follows:
[0013] in The total amount of charging for the energy storage system during the scheduling cycle. This represents the total discharge of the energy storage system during the scheduling cycle. The rated capacity of the energy storage system is given. The multi-objective optimization scheduling module uses the non-dominated sorting genetic algorithm (NSGA-II) with an elite strategy to solve the function. Through population initialization, fast non-dominated sorting, crowding calculation and selection, crossover and mutation operations, Pareto optimal solution set is generated iteratively. Then, the optimal scheduling strategy is selected from the solution set according to the decision requirements. The population size is set to 100, the crossover probability is 0.9, the mutation probability is 0.1, and the number of iterations is 200.
[0014] Preferably, the real-time energy management and scheduling module of the local edge controller (ECC) is based on the optimization strategy issued by the cloud scheduling center, combined with real-time collected elevator operating parameters (including elevator direction of travel, current floor, load, speed v and acceleration a), energy storage system SOC and grid voltage and frequency, etc., and uses the model predictive control (MPC) algorithm for real-time energy allocation. The prediction model of the MPC algorithm is as follows:
[0015] in for The state vector at any given time (including state variables such as common DC bus voltage, energy storage system SOC, and elevator motor power). for Control vectors at any given time (including control quantities such as DC / DC module duty cycle and bidirectional AC / AC converter power command). The state matrix, The control matrix is used; the objective function is:
[0016] in This is a reference value for grid interaction power. This is the SOC reference value for the energy storage system. This is the reference value for the power of the common DC bus. , , , These are the weighting coefficients. To predict the number of steps, the objective function is solved through rolling optimization to obtain the optimal control quantity at the current moment and execute it, thereby achieving real-time and precise energy scheduling.
[0017] Preferably, the bidirectional DC / DC converter of the shared energy storage system adopts a dual closed-loop control strategy of voltage and current. The outer voltage loop is used to maintain the stability of the common DC bus voltage, and its reference voltage is... Based on the optimization strategy of the cloud dispatch center and the current energy interaction requirements between the elevator group and the energy storage system, the output of the voltage outer loop serves as the reference current for the current inner loop. The current inner loop controls the magnitude and direction of the current between the energy storage system and the common DC bus. The current inner loop employs a proportional resonant (PR) controller, whose transfer function is:
[0018] in This is the proportionality coefficient. The resonance coefficient, The fundamental angular frequency is used; through this dual closed-loop control strategy, the voltage fluctuation range of the common DC bus is controlled within ±3% of the rated value, while achieving a rapid response for bidirectional energy flow with a response time of less than 50ms.
[0019] Another technical problem to be solved by this invention is to provide a shared elevator energy-saving and energy storage method based on energy feedback and collaborative scheduling, applied to the system described above, including the following steps: Step 1: Data Acquisition and Upload. The local edge controller (ECC) collects elevator operating status parameters (running direction, floor, load, speed, acceleration), voltage, current, temperature and SOC of each energy storage medium in the shared energy storage system, as well as voltage, frequency and electricity price signals on the grid side through sensors in real time, and uploads the collected real-time data to the cloud dispatch center at a frequency of 100ms. Step Two: Cloud-Edge Collaborative Optimization. The big data analysis and machine learning module of the cloud dispatch center analyzes historical data and uses a Long Short-Term Memory (LSTM) network to predict electricity prices and elevator load for future periods. The input to the LSTM network is the historical electricity price sequence, the historical elevator load sequence, and time characteristics. The output is the predicted electricity price and load for future periods. The loss function of the LSTM network is:
[0020] in This is the actual value. For predicted values, The number of samples is given; the multi-objective optimization scheduling module combines the prediction results with the current system state, solves the multi-objective optimization function using the NSGA-II algorithm, generates an optimized scheduling strategy, and sends it to the local edge controller (ECC). Step 3: Real-time Energy Scheduling and Execution. Based on the optimization strategy issued by the cloud scheduling center, the local edge controller (ECC) uses model predictive control algorithms to perform real-time control of the elevator group's DC / DC modules, the bidirectional DC / DC converters of the shared energy storage system, and the bidirectional AC / AC converters. Specifically: When the elevator is in braking or descending heavy load state, the corresponding DC / DC module is controlled to deliver regenerated power to the common DC bus. If the voltage of the common DC bus is higher than the set value, the bidirectional DC / DC converter is controlled to charge the modular energy storage unit. When the elevator is in an upward heavy load state, it will first obtain power from the common DC bus or modular energy storage unit. If the energy is insufficient, it will draw power from the grid. At the same time, according to the grid electricity price and sharing demand, the bidirectional AC / AC converter is controlled to realize the power interaction between the energy storage system and the grid or to supply power to external sharing users. Step 4: Safety Protection and Monitoring. The Battery Management System (BMS) monitors the status of each energy storage medium in the modular energy storage unit in real time. When the voltage exceeds the rated range by ±5%, the temperature exceeds the normal range (0-55℃ for lithium batteries), or the SOC exceeds the safe range of 0.1-0.9, the charging and discharging circuit of the corresponding energy storage medium is immediately cut off, and an alarm signal is sent to the local edge controller (ECC). The local edge controller (ECC) adjusts the energy dispatch strategy according to the alarm signal to ensure the safe operation of the system.
[0021] Preferably, in the cloud-edge collaborative optimization step, the Long Short-Term Memory (LSTM) network structure includes an input layer, a hidden layer, and an output layer. The input layer dimension is the sum of the time step and the number of features of the historical data. The hidden layer contains 128 LSTM units, and the state update formula for each LSTM unit is:
[0022] in For input gate, For the Gate of Oblivion For output gate, Candidate memory units, For the state of the memory unit, In hidden state, It is the Sigmoid activation function. The hyperbolic tangent activation function is used. , , , This is the weight matrix. , , , The bias vector is used to train the network parameters using the Adam optimizer. The initial learning rate is set to 0.001, and it is reduced to 0.8 every 50 training rounds until the loss function converges, so that the electricity price prediction error is less than 5% and the load prediction error is less than 8%.
[0023] Preferably, in the real-time energy scheduling and execution steps, a time-sharing and zone-based priority scheduling strategy is adopted for the energy sharing management of external users. First, based on the registration information, historical electricity consumption records, and current electricity requests (including power consumption, electricity consumption period, and electricity reliability requirements) of the sharing users, a user priority evaluation model is constructed:
[0024] in Prioritize users Historical telecommunications credit score (values range from 0 to 10, with higher values indicating better credit). The power demand is expressed in kW (normalized value is 0-1). The matching degree between the electricity consumption period and the grid's off-peak period (the matching degree for the off-peak period is 1, for the flat period it is 0.6, and for the peak period it is 0.3). The power reliability requirement is 1 (high reliability is 1, and general reliability is 0.7). , , , The weighting coefficients and + + + =1; according to Users of the shared system are prioritized, with higher-priority users receiving power first. When the power available from the shared energy storage system is insufficient, power is allocated sequentially from highest to lowest priority until the needs of all high-priority users are met or the power allocation is exhausted. Simultaneously, the power consumption status of shared users is monitored in real time. When a user's power consumption exceeds the requested value by more than 10%, the user's subsequent power allocation priority is automatically reduced.
[0025] Preferably, the modular energy storage unit of the shared energy storage system adopts a hybrid energy storage topology, consisting of a supercapacitor bank and a high-nickel battery or solid-state battery bank. The supercapacitor bank is used to smooth out short-term high-frequency fluctuations in the elevator's regenerated power, while the high-nickel battery or solid-state battery bank is used to store long-term low-frequency energy. Both are connected in parallel to a common DC bus through their respective bidirectional DC / DC converters. The energy distribution strategy of the hybrid energy storage is based on wavelet transform, firstly processing the elevator regenerated power signal... Wavelet decomposition is performed to obtain high-frequency components. with low-frequency components The decomposition formula is:
[0026] The high-frequency components correspond to rapidly changing power over short periods of time, which is handled by the supercapacitor bank, i.e., the power command of the supercapacitor bank. The low-frequency components correspond to power that changes slowly over long periods, and are handled by high-nickel batteries or solid-state battery packs, i.e., the power command from high-nickel batteries or solid-state battery packs. This strategy reduces the number of charge-discharge cycles for supercapacitors, extending their lifespan by more than 30%. It also reduces the depth of charge-discharge for high-nickel or solid-state batteries, extending their cycle life by more than 20%. Simultaneously, it improves the absorption efficiency of the entire shared energy storage system for elevator regenerative energy, achieving an absorption efficiency of over 92%.
[0027] This invention provides a shared elevator energy-saving and energy storage system and method based on energy feedback and collaborative scheduling. It has the following beneficial effects: 1. This invention achieves full recovery and efficient utilization of elevator regenerative energy through a "common DC bus + hybrid energy storage" topology. The on-site equipment layer concentrates the regenerated power of the elevator group to a common DC bus, avoiding the dispersed dissipation of regenerated power from individual elevators. The shared energy storage system adopts a hybrid topology of supercapacitor + high-nickel battery or solid-state battery. By decomposing the regenerated power signal through wavelet transform, it solves the problem of short-term high-frequency fluctuations impacting high-nickel batteries or solid-state batteries. The regenerated power recovery efficiency is increased to over 92%, reducing the elevator foundation energy consumption by 25%-35% compared to the traditional resistive dissipation mode.
[0028] The bidirectional DC / DC converter adopts dual closed-loop control of voltage and current to ensure that regenerated power is stably fed into the energy storage system, avoid damage to the elevator frequency converter caused by sudden rise in bus voltage, and extend the life of elevator electrical components by more than 15%.
[0029] 2. This invention constructs a two-layer scheduling architecture of "local edge control + cloud-based global optimization," breaking through the limitations of traditional coarse-grained control: The cloud-based system accurately predicts future electricity prices and elevator loads through an LSTM network, and solves multi-objective optimization functions using the NSGA-II algorithm. This achieves "the highest overall energy efficiency, the lowest grid interaction cost, and the highest energy storage utilization rate," reducing grid interaction costs by 40%-60% compared to the traditional no-optimization mode, and increasing the energy storage system utilization rate from below 30% to over 75%.
[0030] The local edge controller uses the MPC algorithm to respond in real time to changes in elevator operating status and energy storage parameters, dynamically adjusts the duty cycle of the DC / DC module and the power command of the converter, ensures the accuracy of the optimization strategy implementation, and controls the power deviation of the common DC bus within ±2%, avoiding additional energy consumption caused by energy distribution imbalance.
[0031] 3. This invention breaks through the limitations of "exclusive energy storage" and realizes cross-entity sharing and utilization of energy storage resources: Based on a user priority assessment model, external users are ranked according to dimensions such as credit score and load matching degree, and the needs of high-priority users are given priority; during non-office hours in office buildings, idle energy storage can supply power to surrounding shops, and feed back power to the grid during peak hours, realizing "peak-shifting utilization" of energy storage resources, with annual added value of tens of thousands of yuan for a single system; the modular energy storage design supports flexible combination of lithium batteries and supercapacitors, adapting to the energy needs of different scenarios, avoiding resource waste caused by "one-size-fits-all" energy storage configuration, and improving system adaptability by more than 60%.
[0032] 4. This invention constructs a triple security defense line of "hardware + algorithm + strategy" to solve the safety hazards of energy storage and elevator collaborative operation: The battery management system monitors the voltage, temperature, and SOC of the energy storage medium in real time. When the threshold is exceeded, the circuit is immediately cut off and an alarm is triggered. The cycle life of high-nickel batteries or solid-state batteries is extended by more than 20%, and the number of charge and discharge cycles of supercapacitors is reduced by 30%. The local edge controller has a built-in safety protection module. When the grid voltage frequency is abnormal or the elevator malfunctions, it automatically switches to "independent operation mode". Energy storage prioritizes emergency power supply to the elevator to avoid elevator entrapment accidents. The system has a fault-free operation time of ≥10,000 hours. Data transmission adopts an encrypted protocol, and the Pareto optimal solution set of the cloud dispatch center is verified multiple times to avoid misallocation of energy caused by malicious commands and ensure the power safety of the grid and the user side. Attached Figure Description
[0033] Figure 1 This is a schematic diagram of the overall method flow of the present invention. Detailed Implementation
[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0035] Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the invention, and should not be construed as limiting the invention.
[0036] Example 1 A preferred embodiment of the shared elevator energy-saving and energy storage system and method based on energy feedback and collaborative scheduling provided by the present invention is as follows: Figure 1 As shown: A shared elevator energy-saving and energy storage system based on energy feedback and collaborative scheduling includes a local edge controller (ECC), a cloud scheduling center, a field device layer, and a shared energy storage system; The local edge controller (ECC) is used to collect elevator operating status parameters, energy storage system state of charge (SOC) and grid electricity price signals, upload real-time data to the cloud dispatch center, and receive optimization strategies issued by the cloud dispatch center to perform real-time energy management and scheduling, as well as strategy execution and safety protection. The cloud dispatch center is equipped with a big data analysis and machine learning module and a multi-objective optimization dispatch module. The big data analysis and machine learning module learns from historical data on electricity price trends, load patterns and elevator usage patterns. The multi-objective optimization dispatch module aims to achieve the best energy efficiency of the elevator system, the lowest grid interaction cost and the highest utilization rate of shared energy storage. It constructs a multi-objective optimization function and solves it to generate an optimization strategy. The field equipment layer includes an elevator group, frequency converters corresponding to each elevator, and a common DC bus. The frequency converter of each elevator is connected to the common DC bus through a DC / DC module, which is used to transmit the regenerative power generated when the elevator brakes or is under heavy load during descent to the common DC bus, or to obtain power from the common DC bus to drive the elevator to run. The shared energy storage system includes a bidirectional DC / DC converter, modular energy storage units, a battery management system (BMS), and a bidirectional AC / AC converter. The bidirectional DC / DC converter is connected to a common DC bus and is used to regulate the voltage of the common DC bus and realize bidirectional energy flow between the common DC bus and the modular energy storage units. The modular energy storage units are composed of at least one energy storage medium selected from lithium batteries, lead-acid batteries or high-nickel batteries, solid-state batteries, and supercapacitors. The battery management system (BMS) is used to monitor the voltage, current, temperature, and SOC of each energy storage medium and to perform charge and discharge protection. The bidirectional AC / AC converter connects the modular energy storage units to the national power grid and is used to realize bidirectional AC power interaction between the energy storage system and the power grid, while supporting the grid to charge the energy storage system and the energy storage system to feed back power to the grid.
[0037] The multi-objective optimization function constructed by the multi-objective optimization scheduling module of the cloud scheduling center is as follows:
[0038] in The combined energy efficiency function of the elevator system and the energy storage system is expressed as follows:
[0039] in This refers to the amount of regenerated electrical energy recovered by the elevator. The electricity provided by shared energy storage to the outside world Electrical energy input from the power grid, This refers to system energy loss. in The power grid interaction cost function is expressed as follows:
[0040] in for The power exchange with the power grid at any time (positive input, negative output). for The grid electricity price at any given time, where T is the dispatch period; in The shared energy storage utilization rate function is expressed as follows:
[0041] in The total amount of charging for the energy storage system during the scheduling cycle. This represents the total discharge of the energy storage system during the scheduling cycle. The rated capacity of the energy storage system is given. The multi-objective optimization scheduling module uses the non-dominated sorting genetic algorithm (NSGA-II) with an elite strategy to solve the function. Through population initialization, fast non-dominated sorting, crowding calculation and selection, crossover and mutation operations, Pareto optimal solution set is generated iteratively. Then, the optimal scheduling strategy is selected from the solution set according to the decision requirements. The population size is set to 100, the crossover probability is 0.9, the mutation probability is 0.1, and the number of iterations is 200.
[0042] The real-time energy management and scheduling module of the local edge controller (ECC) is based on the optimization strategy issued by the cloud scheduling center. It combines real-time collected elevator operating parameters (including elevator direction, current floor, load, operating speed v, and acceleration a), energy storage system SOC, and grid voltage and frequency parameters, and uses a model predictive control (MPC) algorithm for real-time energy allocation. The prediction model of the MPC algorithm is as follows:
[0043] in for The state vector at any given time (including state variables such as common DC bus voltage, energy storage system SOC, and elevator motor power). for Control vectors at any given time (including control quantities such as DC / DC module duty cycle and bidirectional AC / AC converter power command). The state matrix, The control matrix is used; the objective function is:
[0044] in This is a reference value for grid interaction power. This is the SOC reference value for the energy storage system. This is the reference value for the power of the common DC bus. , , , These are the weighting coefficients. To predict the number of steps, the objective function is solved through rolling optimization to obtain the optimal control quantity at the current moment and execute it, thereby achieving real-time and precise energy scheduling.
[0045] The bidirectional DC / DC converter of the shared energy storage system adopts a dual closed-loop control strategy of voltage and current. The outer voltage loop is used to maintain the stability of the common DC bus voltage, and its reference voltage is... Based on the optimization strategy of the cloud dispatch center and the current energy interaction requirements between the elevator group and the energy storage system, the output of the voltage outer loop serves as the reference current for the current inner loop. The current inner loop controls the magnitude and direction of the current between the energy storage system and the common DC bus. The current inner loop employs a proportional resonant (PR) controller, whose transfer function is:
[0046] in This is the proportionality coefficient. The resonance coefficient, The fundamental angular frequency is used; through this dual closed-loop control strategy, the voltage fluctuation range of the common DC bus is controlled within ±3% of the rated value, while achieving a rapid response for bidirectional energy flow with a response time of less than 50ms.
[0047] Example 2 Please see Figure 1 Furthermore, based on Example 1, a shared elevator energy-saving and energy storage method based on energy feedback and collaborative scheduling is obtained, applied to the system described above, including the following steps: Step 1: Data Acquisition and Upload. The local edge controller (ECC) collects elevator operating status parameters (running direction, floor, load, speed, acceleration), voltage, current, temperature and SOC of each energy storage medium in the shared energy storage system, as well as voltage, frequency and electricity price signals on the grid side through sensors in real time, and uploads the collected real-time data to the cloud dispatch center at a frequency of 100ms. Step Two: Cloud-Edge Collaborative Optimization. The big data analysis and machine learning module of the cloud dispatch center analyzes historical data and uses a Long Short-Term Memory (LSTM) network to predict electricity prices and elevator load for future periods. The input to the LSTM network is the historical electricity price sequence, the historical elevator load sequence, and time characteristics. The output is the predicted electricity price and load for future periods. The loss function of the LSTM network is:
[0048] in This is the actual value. For predicted values, The number of samples is given; the multi-objective optimization scheduling module combines the prediction results with the current system state, solves the multi-objective optimization function using the NSGA-II algorithm, generates an optimized scheduling strategy, and sends it to the local edge controller (ECC). Step 3: Real-time Energy Scheduling and Execution. Based on the optimization strategy issued by the cloud scheduling center, the local edge controller (ECC) uses model predictive control algorithms to perform real-time control of the elevator group's DC / DC modules, the bidirectional DC / DC converters of the shared energy storage system, and the bidirectional AC / AC converters. Specifically: When the elevator is in braking or descending heavy load state, the corresponding DC / DC module is controlled to deliver regenerated power to the common DC bus. If the voltage of the common DC bus is higher than the set value, the bidirectional DC / DC converter is controlled to charge the modular energy storage unit. When the elevator is in an upward heavy load state, it will first obtain power from the common DC bus or modular energy storage unit. If the energy is insufficient, it will draw power from the grid. At the same time, according to the grid electricity price and sharing demand, the bidirectional AC / AC converter is controlled to realize the power interaction between the energy storage system and the grid or to supply power to external sharing users. Step 4: Safety Protection and Monitoring. The Battery Management System (BMS) monitors the status of each energy storage medium in the modular energy storage unit in real time. When the voltage exceeds the rated range by ±5%, the temperature exceeds the normal range (0-55℃ for lithium batteries), or the SOC exceeds the safe range of 0.1-0.9, the charging and discharging circuit of the corresponding energy storage medium is immediately cut off, and an alarm signal is sent to the local edge controller (ECC). The local edge controller (ECC) adjusts the energy dispatch strategy according to the alarm signal to ensure the safe operation of the system.
[0049] In the cloud-edge collaborative optimization step, the Long Short-Term Memory (LSTM) network structure includes an input layer, a hidden layer, and an output layer. The input layer dimension is the sum of the time step and the number of features of the historical data. The hidden layer contains 128 LSTM units, and the state update formula for each LSTM unit is:
[0050] in For input gate, For the Gate of Oblivion For output gate, Candidate memory units, For the state of the memory unit, In hidden state, It is the Sigmoid activation function. The hyperbolic tangent activation function is used. , , , This is the weight matrix. , , , The bias vector is used to train the network parameters using the Adam optimizer. The initial learning rate is set to 0.001, and it is reduced to 0.8 every 50 training rounds until the loss function converges, so that the electricity price prediction error is less than 5% and the load prediction error is less than 8%.
[0051] In the real-time energy scheduling and execution process, the energy sharing management of external users adopts a time-sharing and zone-based priority scheduling strategy. First, based on the registration information, historical electricity consumption records, and current electricity requests (including power consumption, electricity consumption period, and electricity reliability requirements) of the sharing users, a user priority evaluation model is constructed:
[0052] in Prioritize users Historical telecommunications credit score (values range from 0 to 10, with higher values indicating better credit). The power demand is expressed in kW (normalized value is 0-1). The matching degree between the electricity consumption period and the grid's off-peak period (the matching degree for the off-peak period is 1, for the flat period it is 0.6, and for the peak period it is 0.3). The power reliability requirement is 1 (high reliability is 1, and general reliability is 0.7). , , , The weighting coefficients and + + + =1; according to Users of the shared system are prioritized, with higher-priority users receiving power first. When the power available from the shared energy storage system is insufficient, power is allocated sequentially from highest to lowest priority until the needs of all high-priority users are met or the power allocation is exhausted. Simultaneously, the power consumption status of shared users is monitored in real time. When a user's power consumption exceeds the requested value by more than 10%, the user's subsequent power allocation priority is automatically reduced.
[0053] According to the system or method described above, the modular energy storage unit of the shared energy storage system adopts a hybrid energy storage topology, consisting of a supercapacitor bank and a high-nickel battery or solid-state battery bank. The supercapacitor bank is used to smooth out short-term high-frequency fluctuations in elevator regenerated power, while the high-nickel battery or solid-state battery bank is used to store long-term low-frequency energy. Both are connected in parallel to a common DC bus through their respective bidirectional DC / DC converters. The energy distribution strategy of the hybrid energy storage is based on wavelet transform, firstly by processing the elevator regenerated power signal. Wavelet decomposition is performed to obtain high-frequency components. with low-frequency components The decomposition formula is:
[0054] The high-frequency components correspond to rapidly changing power over short periods of time, which is handled by the supercapacitor bank, i.e., the power command of the supercapacitor bank. The low-frequency components correspond to power that changes slowly over long periods, and are handled by high-nickel batteries or solid-state battery packs, i.e., the power command from high-nickel batteries or solid-state battery packs. This strategy reduces the number of charge-discharge cycles for supercapacitors, extending their lifespan by more than 30%. It also reduces the depth of charge-discharge for high-nickel or solid-state batteries, extending their cycle life by more than 20%. Simultaneously, it improves the absorption efficiency of the entire shared energy storage system for elevator regenerative energy, achieving an absorption efficiency of over 92%.
[0055] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0056] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. An energy feedback and cooperative scheduling based shared elevator energy saving and storage system, characterized in that, The local edge controller, the cloud scheduling center, the field device layer, and the shared energy storage system are included. The local edge controller is used for collecting elevator operation state parameters, energy storage system state of charge, and power grid price signals, uploading real-time data to the cloud scheduling center, receiving optimization strategies issued by the cloud scheduling center, and performing real-time energy management and scheduling and strategy execution and safety protection. The cloud scheduling center is configured with a big data analysis and machine learning module and a multi-objective optimization scheduling module. The big data analysis and machine learning module learns the price trend, load law, and elevator usage mode based on historical data. The multi-objective optimization scheduling module takes the elevator system energy efficiency optimization, the lowest grid interaction cost, and the maximum shared energy storage utilization rate as the target, constructs a multi-objective optimization function, and solves it to generate an optimization strategy. The field device layer includes an elevator group, a frequency converter corresponding to each elevator, and a public DC bus. The frequency converter of each elevator is connected to the public DC bus through a DC / DC module, used to deliver regenerated electric energy generated when the elevator brakes or descends to the public DC bus, or obtain electric energy from the public DC bus to drive the elevator to run. The shared energy storage system includes a bidirectional DC / DC converter, a modular energy storage unit, a battery management system, and a bidirectional AC / AC converter. The bidirectional DC / DC converter is connected to the public DC bus, used to adjust the public DC bus voltage and realize the bidirectional flow of energy between the public DC bus and the modular energy storage unit. The modular energy storage unit is composed of at least one energy storage medium selected from lithium batteries, lead-acid batteries, or high-nickel batteries, solid-state batteries, and super capacitors. The battery management system is used to monitor the voltage, current, temperature, and SOC of each energy storage medium and perform charge and discharge protection. The bidirectional AC / AC converter connects the modular energy storage unit with the national power grid, used to realize the bidirectional interaction of alternating current energy between the energy storage system and the power grid, while supporting the charging of the energy storage system by the power grid and the feedback of electric energy from the energy storage system to the power grid.
2. The system of claim 1, wherein, The multi-objective optimization function constructed by the multi-objective optimization scheduling module of the cloud scheduling center is: ; wherein is a comprehensive energy efficiency function of the elevator system and the energy storage system, and the expression is ; wherein is the elevator regenerative energy recovery amount, is the shared energy storage provided to the outside, is the power input from the grid, is the system energy loss; wherein is the grid interaction cost function, expressed as ; wherein is the interaction power with the grid at the moment (positive for input, negative for output), is the grid price at the moment, T is the dispatch period; wherein is the shared energy storage utilization rate function, expressed as ; wherein is the total charging amount of the energy storage system in the dispatching period, is the total discharging amount of the energy storage system in the dispatching period, is the rated capacity of the energy storage system; the multi-objective optimization scheduling module solves the function by using a non-dominated sorting genetic algorithm II (NSGA-II) with an elite strategy, generates a Pareto optimal solution set through initialization of a population, fast non-dominated sorting, calculation and selection of crowding degree, crossover, and mutation operations, and iterates to select an optimal scheduling strategy from the solution set according to decision requirements, wherein the population size is set to 100, the crossover probability is 0.9, the mutation probability is 0.1, and the number of iterations is 200 times.
3. The system of claim 1, wherein, The real-time energy management and scheduling module of the local edge controller uses a model predictive control algorithm to perform real-time energy distribution based on the optimization strategy issued by the cloud scheduling center and the real-time collected elevator operation parameters, energy storage system SOC, and power grid voltage frequency parameters. The prediction model of the MPC algorithm is: ; where is the state vector at time t, is the control vector at time t, is the state matrix, is the control matrix; and the objective function is: ; wherein is a grid interaction power reference value, is an energy storage system SOC reference value, is a common DC bus power reference value, , , , is a weighting factor, is a prediction step number.
4. The system of claim 1, wherein, The bidirectional DC / DC converter of the shared energy storage system adopts a voltage and current double-loop control strategy. The voltage outer loop is used to maintain the stability of the common DC bus voltage, and the reference voltage According to the optimization strategy of the cloud scheduling center and the energy interaction demand setting of the current elevator group and the energy storage system, the output of the voltage outer loop is used as the reference current of the current inner loop. The current inner loop is used to control the current size and direction between the energy storage system and the common DC bus. The current inner loop adopts a proportional resonant (PR) controller, and the transfer function is: ; wherein is a proportionality factor, is a resonance factor, is the fundamental angular frequency.
5. A shared elevator energy saving and energy storage method based on energy feedback and cooperative scheduling, applied to the system of any one of claims 1-4, characterized in that, The method comprises the following steps: Step one, data collection and uploading step, the local edge controller collects elevator operation state parameters, voltage, current, temperature, and SOC of each energy storage medium of the shared energy storage system, and voltage, frequency, and price signals on the power grid side in real time through sensors, and uploads the collected real-time data to the cloud scheduling center at a frequency of 100 ms; Step two, cloud-edge collaborative optimization step, the big data analysis and machine learning module of the cloud scheduling center analyzes historical data, uses a long short-term memory network to predict the price and elevator usage load in the future period, the input of the LSTM network is the historical price sequence, historical elevator operation load sequence, and time characteristics, and the output is the predicted value of the price and load in the future period, and the loss function of the LSTM network is: ; wherein is the actual value, is the predicted value, is the number of samples; the multi-objective optimization scheduling module combines the prediction result and the current system state, solves a multi-objective optimization function through an NSGA-II algorithm, generates an optimized scheduling strategy, and issues the optimized scheduling strategy to a local edge controller; Step three, real-time energy scheduling and execution step, based on the optimization strategy issued by the cloud scheduling center, the local edge controller uses model predictive control algorithm to control the DC / DC module of the elevator group, the bidirectional DC / DC converter and the bidirectional AC / AC converter of the shared energy storage system in real time, specifically: When the elevator is in the braking or down heavy load state, control the corresponding DC / DC module to deliver regenerative electric energy to the public DC bus, if the public DC bus voltage is higher than the set value, control the bidirectional DC / DC converter to charge the modular energy storage unit; When the elevator is in the up heavy load state, preferentially obtain electric energy from the public DC bus or the modular energy storage unit, if the energy is insufficient, take power from the power grid; At the same time, according to the power price and the sharing demand, control the bidirectional AC / AC converter to realize the energy exchange between the energy storage system and the power grid or supply power to the external sharing user; Step four, safety protection and monitoring step, the battery management system monitors the state of each energy storage medium of the modular energy storage unit in real time, when it is detected that the voltage exceeds the rated range ± 5%, the temperature exceeds the normal range or the SOC exceeds the safety interval of 0.1-0.9, immediately cut off the charge and discharge circuit of the corresponding energy storage medium, and send an alarm signal to the local edge controller, the local edge controller adjusts the energy scheduling strategy according to the alarm signal to ensure the safe operation of the system.
6. The method of claim 5, wherein, In the cloud edge collaborative optimization step, the network structure of the long short-term memory network includes input layer, hidden layer and output layer, the dimension of the input layer is the sum of the time step and the feature number of the historical data, the hidden layer contains 128 LSTM units, and the state update formula of each LSTM unit is: ; wherein is an input gate, is a forget gate, is an output gate, is a candidate memory cell, is a memory cell state, is a hidden state, is a sigmoid activation function, is a tanh activation function, , , , is a weight matrix, , , , is a bias vector.
7. The method of claim 5, wherein, In the real-time energy scheduling and execution step, the time-sharing and zoning priority scheduling strategy is adopted for the electric energy sharing management of the external sharing user, first, according to the registration information, historical power consumption record and current power consumption request of the sharing user, a user priority evaluation model is constructed: ; wherein is a user priority, is a historical electricity usage credit score, is an electricity power demand, is a match degree of electricity usage time period and grid valley time period, is an electricity reliability requirement, , , , is a weight coefficient and + + + = 1. According to The shared users are prioritized, and the users with high priority are given the power supply first. When the shared energy storage system has insufficient power supply, the power is allocated in order of priority from high to low until all high-priority users' demands are met or the power is allocated completely. Meanwhile, the power consumption state of the shared users is monitored in real time, and when the power consumption of a user exceeds the request value by more than 10%, the subsequent power allocation priority of the user is automatically reduced.
8. The system or method of any of claims 1-7, wherein, The modular energy storage unit of the shared energy storage system adopts a hybrid energy storage topology, which is composed of a super capacitor group and a high-nickel battery or a solid-state battery group, the super capacitor group is used to suppress short-time high-frequency fluctuations of elevator regenerative power, the high-nickel battery or the solid-state battery group is used to store long-time low-frequency energy, and the two are connected in parallel on a common DC bus through respective bidirectional DC / DC converters; the energy distribution strategy of the hybrid energy storage is realized based on wavelet transform, first, wavelet decomposition is performed on the elevator regenerative power signal to obtain high-frequency components and low-frequency components , and the decomposition formula is: ; wherein the high frequency component corresponds to short-time fast changing power, which is borne by the super capacitor pack, i.e. the power instruction of the super capacitor pack , the low frequency component corresponds to long-time slow changing power, which is borne by the high nickel battery or solid-state battery pack, i.e. the power instruction of the high nickel battery or solid-state battery pack .