A big data-based electric energy storage and calling control method and system

Abstract

The application provides a kind of based on big data's electric energy storage and calling control method and system, it is related to electric power automation technical field, the method includes: obtaining elevator cabinet control signal and power grid state information;Obtain elevator operating state information;According to elevator operating state information, elevator cabinet control signal and power grid state information, determine electric energy recovery discrimination parameter;Determine whether to carry out electric energy recovery;In the case where electric energy recovery is needed, the stored electric energy is recovered;Real-time date data, real-time weather data, building event arrangement data and real-time monitoring image are obtained;According to the trained passenger flow prediction model, elevator operating state information, real-time date data, real-time weather data, building event arrangement data and real-time monitoring image are processed to determine the predicted passenger flow;Determine the electric energy distribution coefficient;According to electric energy distribution coefficient, carry out electric energy calling control.According to the application, the effectiveness of electric energy storage and calling control can be improved.

Description

A method and system for power storage and retrieval control based on big data Technical Field

[0001] This invention relates to the field of power automation technology, and in particular to a method and system for power storage and retrieval control based on big data. Background Technology

[0002] In related technologies, power can be temporarily allocated and controlled based on passenger flow. However, this temporary allocation and control may result in multiple elevators competing for peak power resources simultaneously, which could impact the power grid and make the system difficult to operate stably. In other words, it is difficult to allocate and allocate power to elevators in advance to improve the stability of the system.

[0003] The information disclosed in the background section of this application is intended only to enhance the understanding of the general background of this application and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0004] This invention provides a method and system for energy storage and retrieval control based on big data, which can solve the technical problem that related technologies are unable to pre-allocate and retrieval energy for elevators and improve the stability of system operation.

[0005] According to a first aspect of the present invention, a method for energy storage and retrieval control based on big data is provided, comprising:

[0006] During the control cycle, acquire elevator cabinet control signals and power grid status information;

[0007] During the control cycle, acquire elevator operating status information;

[0008] Based on the elevator operating status information, the elevator cabinet control signal, and the power grid status information, the energy recovery discrimination parameters are determined;

[0009] Based on the energy recovery discrimination parameters, determine whether to perform energy recovery;

[0010] When energy recovery is required, the recovered energy is stored.

[0011] Acquire real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images;

[0012] The elevator operating status information, real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images are processed based on the trained passenger flow prediction model to determine the predicted passenger flow.

[0013] Based on the predicted passenger flow, determine the power allocation coefficient;

[0014] Power dispatch control is performed based on the power allocation coefficient.

[0015] According to the present invention, energy recovery discrimination parameters are determined based on the elevator operating status information, the elevator cabinet control signal, and the power grid status information, including:

[0016] The first recovery judgment result is determined based on the elevator cabinet control signal;

[0017] Based on the grid status information, determine the DC bus voltage, grid current harmonics, power factor, grid voltage fluctuation, frequency deviation, and battery capacity.

[0018] The second recycling discrimination result is determined based on the DC bus voltage, the grid current harmonics, the power factor, the grid voltage fluctuation, the frequency deviation, and the battery capacity.

[0019] Based on the elevator operating status information, determine the elevator mass, elevator load, and elevator descent height;

[0020] The third recovery judgment result is determined based on the elevator mass, the elevator load, and the elevator descent height;

[0021] Based on the first recovery discrimination result, the second recovery discrimination result, and the third recovery discrimination result, the energy recovery discrimination parameters are determined.

[0022] According to the present invention, a second recycling discrimination result is determined based on the DC bus voltage, the grid current harmonics, the power factor, the grid voltage fluctuation, the frequency deviation, and the battery capacity, including:

[0023] The voltage and current constraint results are determined based on the DC bus voltage, the grid-connected current harmonics, and the power factor.

[0024] The power grid state constraint results are determined based on the power grid voltage fluctuations and the frequency deviations.

[0025] Based on the battery capacity, determine the battery state constraint result;

[0026] Based on the voltage and current constraint results, the power grid state constraint results, and the battery state constraint results, a second recovery discrimination result is determined.

[0027] According to the present invention, a third recovery judgment result is determined based on the elevator mass, the elevator load, and the elevator descent height, including:

[0028] The estimated recovery power is determined based on the elevator mass, the elevator load, and the elevator descent height.

[0029] Obtain the rated power of the elevator motor;

[0030] Determine the preset recovery power threshold;

[0031] The third recovery discrimination result is determined based on the expected recovery power, the rated power of the elevator motor, and the preset recovery power threshold.

[0032] According to the present invention, the training steps of the passenger flow prediction model include:

[0033] Acquire historical passenger flow, elevator operation status information, historical date data, historical weather data, historical building event scheduling data, and historical monitoring images for historical control cycles;

[0034] Based on the historical elevator operation status information, determine the proportion of historically operable elevators and the historical operable identification results;

[0035] Based on the historical date data and the historical passenger flow, the historical workday identification result is determined;

[0036] Based on the historical weather data, the historical weather severity coefficient is determined, wherein the historical weather data includes: historical temperature data, historical wind speed data, and historical rainfall data;

[0037] Based on the historical building event scheduling data, determine the number of participants in the historical building events corresponding to multiple moments in the historical control cycle;

[0038] Based on the historical surveillance images, determine the number of people queuing in the past;

[0039] The passenger flow prediction model is used to process the historical operational identification results, the historical operational elevator ratio, the historical weekday identification results, the historical weather severity coefficient, the number of participants in historical building events, and the number of people queuing in the past to obtain the predicted sample passenger flow.

[0040] The training loss function of the passenger flow prediction model is determined based on the historical passenger flow, the predicted sample passenger flow, the historical proportion of operational elevators, the historical operational identification results, the historical weekday identification results, the historical weather severity coefficient, the number of participants in historical building events, and the historical number of people queuing.

[0041] Based on the training loss function of the passenger flow prediction model, the passenger flow prediction model is trained to obtain the trained passenger flow prediction model.

[0042] According to the present invention, determining the historical weather severity coefficient based on the historical weather data includes: according to the formula:

[0043]

[0044] Determine the historical weather severity coefficient at time j in the i-th historical control cycle. ,in, This provides the historical temperature data for the j-th moment of the i-th historical control cycle. To preset historical temperature data thresholds, This provides the historical wind speed data for the j-th moment of the i-th historical control cycle. To preset historical wind speed data thresholds, For the historical rainfall data at time j in the i-th historical control cycle, This is a preset threshold for historical rainfall data.

[0045] According to the present invention, the training loss function of the passenger flow prediction model is determined based on the historical passenger flow, the predicted sample passenger flow, the historical proportion of operational elevators, the historical operational identification results, the historical weekday identification results, the historical weather severity coefficient, the number of participants in historical building events, and the historical queuing length, including: according to the formula:

[0046]

[0047] Determine the training loss function for the passenger flow prediction model. ,in, Let the predicted sample passenger flow of the k-th elevator be the passenger flow at the j-th time point in the i-th historical control cycle. Let K be the historical passenger flow of elevator K during the period from time j in the i-th historical control cycle to five minutes after time j. This represents the proportion of historically operable elevators at time j in the i-th historical control cycle. To preset the threshold for the proportion of elevators that can operate, This represents the historical operational availability identification result of the k-th elevator at the j-th moment of the i-th historical control cycle. , Let be the historical weather severity coefficient at time j in the i-th historical control cycle. To preset the historical weather severity coefficient threshold, Let be the historical queue size at time j in the i-th historical control cycle. To preset a historical threshold for the number of people in the queue, The number of participants in the historical building event at time j of the i-th historical control period. To preset a threshold for the number of participants in historical building events, The result of identifying the historical working day for the i-th historical control cycle is given, where n is the number of historical control cycles (i ≤ n), j is the number of moments in the historical control cycle (j ≤ m), and i, n, j, and m are all positive integers.

[0048] According to a second aspect of the present invention, a big data-based power storage and retrieval control system is provided, comprising:

[0049] The integrated information module is used to acquire elevator cabinet control signals and power grid status information during the control cycle;

[0050] The operation information module is used to acquire elevator operation status information during the control cycle;

[0051] The discrimination parameter module is used to determine the energy recovery discrimination parameters based on the elevator operating status information, the elevator cabinet control signal, and the power grid status information.

[0052] The energy recovery discrimination module is used to determine whether to perform energy recovery based on the energy recovery discrimination parameters.

[0053] An energy recovery module is used to store recovered energy when it is necessary to recover energy.

[0054] The real-time data module is used to acquire real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images;

[0055] The passenger flow prediction module is used to process the elevator operation status information, the real-time date data, the real-time weather data, the building event scheduling data, and the real-time monitoring images based on the trained passenger flow prediction model to determine the predicted passenger flow.

[0056] The allocation coefficient module is used to determine the power allocation coefficient based on the predicted passenger flow.

[0057] The control module is invoked to control the energy allocation based on the energy distribution coefficient.

[0058] Technical Effects: According to this invention, based on elevator control signals, elevator operating status information, and power grid status information, the current elevator and power grid conditions can be accurately analyzed to determine whether energy recovery is possible. Energy recovery criteria are then determined. If energy recovery is possible, the recovered energy is stored. Furthermore, a passenger flow prediction model can be used to predict the elevator's passenger flow for the next five minutes, and the effectiveness of energy storage and retrieval control is determined based on the predicted passenger flow. When determining historical weather severity coefficients, historical weather data can be used. During the calculation process, weather severity can be assessed based on temperature, wind speed, and rainfall, improving the comprehensiveness and accuracy of historical weather severity coefficients. When determining the training loss function for the passenger flow prediction model, it can be based on historical passenger flow, predicted sample passenger flow, historical proportion of operational elevators, historical operational identification results, historical weekday identification results, historical weather severity coefficient, number of participants in historical building events, and historical queuing numbers. During the calculation process, the potential impact of the proportion of operational elevators, weekday identification results, weather severity coefficient, number of participants in building events, and queuing numbers on the predicted passenger flow can be considered to determine the influence of the above data on the error of the predicted sample passenger flow. Based on this influence and the relative error of the predicted sample passenger flow, the training loss function can be set to reduce the training loss function during the training process, thereby more effectively improving the accuracy of the passenger flow prediction model.

[0059] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Other features and aspects of the invention will become clearer from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0060] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these drawings without creative effort.

[0061] Figure 1 illustrates, by way of example, a flowchart of a big data-based power storage and retrieval control method according to an embodiment of the present invention;

[0062] Figure 2 illustrates, by way of example, a schematic diagram of determining energy recovery discrimination parameters according to an embodiment of the present invention;

[0063] Figure 3 illustrates, exemplarily, a block diagram of a big data-based power storage and retrieval control system according to an embodiment of the present invention. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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.

[0065] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0066] Figure 1 illustrates, by way of example, a flowchart of a big data-based power storage and retrieval control method according to an embodiment of the present invention, the method comprising:

[0067] Step S1: During the control cycle, acquire the elevator cabinet control signal and power grid status information;

[0068] Step S2: During the control cycle, acquire elevator operating status information;

[0069] Step S3: Determine the energy recovery discrimination parameters based on the elevator operating status information, the elevator cabinet control signal, and the power grid status information;

[0070] Step S4: Determine whether to perform energy recovery based on the energy recovery discrimination parameters;

[0071] Step S5: If energy recovery is required, store the recovered energy.

[0072] Step S6: Obtain real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images;

[0073] Step S7: Process the elevator operation status information, real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images according to the trained passenger flow prediction model to determine the predicted passenger flow.

[0074] Step S8: Determine the power allocation coefficient based on the predicted passenger flow;

[0075] Step S9: Perform power dispatch control according to the power allocation coefficient.

[0076] According to an embodiment of the present invention, the energy storage and retrieval control method based on big data can accurately analyze whether the current elevator and power grid conditions allow for energy recovery based on elevator control signals, elevator operating status information, and power grid status information. It determines energy recovery discrimination parameters, stores the recovered energy when energy recovery is possible, and predicts the elevator passenger flow for the next five minutes using a passenger flow prediction model. Based on the predicted passenger flow, it determines whether to retrieval control of the recovered energy, thereby improving the effectiveness of energy storage and retrieval control.

[0077] According to one embodiment of the present invention, in step S1, during the control cycle, the elevator cabinet control signal and power grid status information are acquired.

[0078] For example, elevator control signals (signals displaying elevator braking status, running direction, and speed) can be obtained through the communication port of the elevator control cabinet, and power grid status information (such as DC bus voltage and grid voltage) can be obtained through smart meters or sensors (such as voltage sensors and current sensors).

[0079] According to one embodiment of the present invention, in step S2, elevator operating status information is acquired during the control cycle.

[0080] For example, elevator operating status information (such as elevator mass, elevator load, and elevator descent height) can be obtained through technical documents and sensors installed in the elevator, such as pressure sensors and encoders.

[0081] According to an embodiment of the present invention, in step S3, energy recovery discrimination parameters are determined based on the elevator operating status information, the elevator cabinet control signal, and the power grid status information.

[0082] Figure 2 illustrates, by way of example, a schematic diagram of determining energy recovery discrimination parameters according to an embodiment of the present invention.

[0083] According to an embodiment of the present invention, step S3 includes:

[0084] Step S31: Determine the first recovery judgment result based on the elevator cabinet control signal;

[0085] Step S32: Based on the grid status information, determine the DC bus voltage, grid current harmonics, power factor, grid voltage fluctuation, frequency deviation, and battery capacity.

[0086] Step S33: Determine the second recycling discrimination result based on the DC bus voltage, the grid current harmonics, the power factor, the grid voltage fluctuation, the frequency deviation, and the battery capacity;

[0087] Step S34: Determine the elevator mass, elevator load, and elevator descent height based on the elevator operating status information;

[0088] Step S35: Determine the third recovery judgment result based on the elevator mass, the elevator load, and the elevator descent height;

[0089] Step S36: Determine the energy recovery discrimination parameters based on the first recovery discrimination result, the second recovery discrimination result, and the third recovery discrimination result.

[0090] For example, if an elevator control signal indicating "braking," "descending," and "speed greater than 0.5 m / s" is received, it means the elevator is in a state where it can generate electricity externally, and the first recovery judgment result is 1; otherwise, the first recovery judgment result is 0. The DC bus voltage is directly obtained through a voltage sensor installed across the DC bus capacitor of the inverter. The grid-connected current harmonics are obtained through a current sensor and digital signal processor. The power factor is calculated through a DSP. The grid voltage is continuously monitored through a voltage sensor, and grid voltage fluctuations and frequency deviations are calculated through a DSP. The battery capacity is obtained through a battery management system. Based on the DC bus voltage, The second recovery judgment result is determined by analyzing grid current harmonics, power factor, grid voltage fluctuations, frequency deviation, and battery capacity to determine whether the grid and battery conditions support energy recovery. The elevator mass, load, and descent height are obtained using technical documents and pressure sensors and encoders installed in the elevator. Based on these parameters, the third recovery judgment result is determined to assess whether the energy recovered during the descent is suitable for recovery. When all three recovery judgment results are 1, the energy recovery judgment parameter is 1, indicating that recovery is possible; otherwise, the parameter is 0.

[0091] According to an embodiment of the present invention, step S33 includes:

[0092] Step S331: Determine the voltage and current constraint results based on the DC bus voltage, the grid-connected current harmonics, and the power factor;

[0093] Step S332: Determine the grid state constraint result based on the grid voltage fluctuation and the frequency deviation;

[0094] Step S333: Determine the battery state constraint result based on the battery capacity;

[0095] Step S334: Determine the second recycling discrimination result based on the voltage and current constraint result, the power grid state constraint result, and the battery state constraint result.

[0096] For example, when the DC bus voltage is less than or equal to 1.15 times the rated voltage, the grid current harmonics are less than 3%, and the power factor is greater than or equal to 0.9, it indicates that the system is not over-voltage and the power quality meets the recovery conditions, with a voltage and current constraint result of 1; otherwise, the voltage and current constraint result is 0. When the grid voltage fluctuation is greater than 10% or the frequency deviation is less than 0.5Hz, it indicates that the grid is in a severely abnormal state. In this state, the energy recovery system must stop feeding power back to the grid and switch to resistance consumption mode, with a grid state constraint result of 0; otherwise, the grid state constraint result is 1. When the battery capacity is less than or equal to 95%, it indicates that the possibility of overcharging is small, with a battery state constraint result of 1; otherwise, it indicates that the possibility of overcharging is large, with a battery state constraint result of 0. If the voltage and current constraint result, the grid state constraint result, and the battery state constraint result are all 1, the second recovery judgment result is 1; otherwise, the second recovery judgment result is 0.

[0097] According to an embodiment of the present invention, step S35 includes:

[0098] Step S351: Determine the expected recovery power based on the elevator mass, the elevator load, and the elevator descent height;

[0099] Step S352: Obtain the rated power of the elevator motor;

[0100] Step S353: Determine the preset recovery power threshold;

[0101] Step S354: Determine the third recovery discrimination result based on the expected recovery power, the rated power of the elevator motor, and the preset recovery power threshold.

[0102] For example, the total mass of the elevator is determined based on its mass and load. The expected recovery power is determined based on the total mass and the elevator's descent height. The rated power of the elevator motor is obtained from the elevator's technical documents. The preset recovery power threshold is set to 1 kW. If the recovered electrical energy is less than the preset recovery power threshold (low recovery efficiency) or greater than 80% of the elevator motor's rated power (overvoltage damages the converter), the third recovery judgment result is 0. Otherwise, the third recovery judgment result is 1.

[0103] According to one embodiment of the present invention, in step S4, it is determined whether to perform energy recovery based on the energy recovery discrimination parameters.

[0104] For example, when the energy recovery discrimination parameter is 1, it means that the condition of the power grid, battery and energy meets the conditions for recovery, and energy recovery is determined to be carried out; otherwise, energy recovery is not carried out.

[0105] According to one embodiment of the present invention, in step S5, if energy recovery is required, the recovered energy is stored.

[0106] For example, the AC power generated by the elevator is converted into DC power by a frequency converter. The active front end or PWM rectifier in the system (used as an inverter at this time) inverts the stable DC power into AC power with the same frequency, phase and voltage as the power grid. After being filtered by a reactor, the clean AC power is fed back into the power grid inside the building.

[0107] According to one embodiment of the present invention, in step S6, real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images are acquired.

[0108] For example, real-time date and weather data can be obtained through calendars and weather forecasts, real-time monitoring images in front of elevators can be captured through cameras, and building event schedules can be queried through the building management system (e.g., meeting A has 200 participants and ends at 10 o'clock) to obtain building event schedule data.

[0109] According to an embodiment of the present invention, in step S7, the elevator operating status information, the real-time date data, the real-time weather data, the building event scheduling data, and the real-time monitoring images are processed according to the trained passenger flow prediction model to determine the predicted passenger flow.

[0110] For example, based on a trained passenger flow prediction model, elevator operating status information, real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images can be processed to determine the number of people transported by the elevator at the current moment and within five minutes after the current moment, that is, to predict passenger flow.

[0111] According to an embodiment of the present invention, the training steps of the passenger flow prediction model include:

[0112] Acquire historical passenger flow, elevator operation status information, historical date data, historical weather data, historical building event scheduling data, and historical monitoring images for historical control cycles;

[0113] Based on the historical elevator operation status information, determine the proportion of historically operable elevators and the historical operable identification results;

[0114] Based on the historical date data and the historical passenger flow, the historical workday identification result is determined;

[0115] Based on the historical weather data, the historical weather severity coefficient is determined, wherein the historical weather data includes: historical temperature data, historical wind speed data, and historical rainfall data;

[0116] Based on the historical building event scheduling data, determine the number of participants in the historical building events corresponding to multiple moments in the historical control cycle;

[0117] Based on the historical surveillance images, determine the number of people queuing in the past;

[0118] The passenger flow prediction model is used to process the historical operational identification results, the historical operational elevator ratio, the historical weekday identification results, the historical weather severity coefficient, the number of participants in historical building events, and the number of people queuing in the past to obtain the predicted sample passenger flow.

[0119] The training loss function of the passenger flow prediction model is determined based on the historical passenger flow, the predicted sample passenger flow, the historical proportion of operational elevators, the historical operational identification results, the historical weekday identification results, the historical weather severity coefficient, the number of participants in historical building events, and the historical number of people queuing.

[0120] Based on the training loss function of the passenger flow prediction model, the passenger flow prediction model is trained to obtain the trained passenger flow prediction model.

[0121] For example, the elevator management system obtains historical passenger flow data and historical elevator operating status information for each elevator during historical control periods; the building management system obtains historical building event scheduling data, historical monitoring images, historical date data, and historical weather data; based on the historical elevator operating status information, it determines whether the elevator is in normal operating condition. If the elevator is in normal operating condition, its historical operational status identification result is 1; otherwise, it is 0. The historical operational elevator ratio is determined by the ratio of the sum of the historical operational status identification results of all elevators to the total number of elevators; based on the historical date data... According to the data, it is determined whether the date of the historical control cycle falls on a working day. If it does not, the historical working day identification result is 1. If it does, the historical working day identification result is determined by the ratio of the average historical passenger flow of all elevators over multiple working days to the average historical passenger flow of all elevators over multiple rest and holiday periods. Based on historical temperature data, historical wind speed data, and historical rainfall data, the weather severity of the date of the historical control cycle is assessed to determine the historical weather severity coefficient. Based on historical building event scheduling data, the number of participants in historical building events corresponding to multiple moments in the historical control cycle is determined. For example, when the historical control cycle... If the start time is 10:00, the number of participants in historical building events corresponding to the start time of the historical control cycle is determined by querying the building management system for the total number of participants in all meetings and activities ending between 10:00 and 10:05. The image detection model is used to identify the number of people queuing at elevator entrances, i.e., the historical queue size, based on images from elevator entrances. The image detection model is a type of neural network model, trained using historical data to enable it to recognize people in images. The system also includes historical operational elevator identification results, historical operational elevator ratios, historical weekday identification results, and historical weather severity coefficients, all based on passenger flow prediction models. The process involves processing historical building event participants and historical queue lengths to obtain elevator passenger flow data for multiple moments within a historical control cycle, as well as the passenger flow five minutes after each moment; this is used to predict sample passenger flow. Based on historical passenger flow, predicted sample passenger flow, historical operational elevator ratio, historical operational elevator identification results, historical weekday identification results, historical weather severity coefficient, historical building event participants, and historical queue lengths, the training loss function for the passenger flow prediction model is determined. The passenger flow prediction model is then trained using this training loss function to improve its accuracy, resulting in a trained passenger flow prediction model.

[0122] According to one embodiment of the present invention, determining the historical weather severity coefficient based on the historical weather data includes: determining the historical weather severity coefficient at the j-th moment of the i-th historical control cycle according to formula (1). ,

[0123] (1)

[0124] in, This provides the historical temperature data for the j-th moment of the i-th historical control cycle. To preset historical temperature data thresholds, This provides the historical wind speed data for the j-th moment of the i-th historical control cycle. To preset historical wind speed data thresholds, For the historical rainfall data at time j in the i-th historical control cycle, This is a preset threshold for historical rainfall data.

[0125] According to one embodiment of the present invention, This is the relative difference between the historical temperature data at time j in the i-th historical control cycle and the preset historical temperature data threshold. The larger this ratio, the more likely the historical temperature data is too high or too low, indicating that the weather may be in a high-temperature or low-temperature condition. The preset historical temperature data threshold can be set to 20 degrees Celsius. The ratio is the relative difference between the historical wind speed data at time j in the i-th historical control cycle and the preset historical wind speed data threshold. The larger this ratio, the larger the relative historical wind speed data, indicating that the weather may be in a strong wind condition. The preset historical wind speed data threshold can be set to 8 m / s. The ratio is the relative difference between the historical rainfall data at time j in the i-th historical control cycle and the preset historical rainfall data threshold. The larger the ratio, the larger the historical rainfall data, indicating that the weather may be in a rainstorm condition. The historical rainfall data threshold can be set to 10 mm.

[0126] According to one embodiment of the present invention, This indicates that the historical weather severity coefficient is determined based on three factors: temperature, wind speed, and rainfall.

[0127] In this way, historical weather severity coefficients can be determined based on historical weather data. During the calculation process, the severity of the weather can be assessed based on three aspects: temperature, wind speed, and rainfall, thus improving the comprehensiveness and accuracy of historical weather severity coefficients.

[0128] According to one embodiment of the present invention, the training loss function of the passenger flow prediction model is determined based on the historical passenger flow, the predicted sample passenger flow, the historical proportion of operational elevators, the historical operational identification results, the historical weekday identification results, the historical weather severity coefficient, the number of participants in historical building events, and the historical queuing number, including: determining the training loss function of the passenger flow prediction model according to formula (2). ,

[0129] (2)

[0130] in, Let the predicted sample passenger flow of the k-th elevator be the passenger flow at the j-th time point in the i-th historical control cycle. Let K be the historical passenger flow of elevator K during the period from time j in the i-th historical control cycle to five minutes after time j. This represents the proportion of historically operable elevators at time j in the i-th historical control cycle. To preset the threshold for the proportion of elevators that can operate, This represents the historical operational availability identification result of the k-th elevator at the j-th moment of the i-th historical control cycle. , Let be the historical weather severity coefficient at time j in the i-th historical control cycle. To preset the historical weather severity coefficient threshold, Let be the historical queue size at time j in the i-th historical control cycle. To preset a historical threshold for the number of people in the queue, The number of participants in the historical building event at time j of the i-th historical control period. To preset a threshold for the number of participants in historical building events, The result of identifying the historical working day for the i-th historical control cycle is given, where n is the number of historical control cycles (i ≤ n), j is the number of moments in the historical control cycle (j ≤ m), and i, n, j, and m are all positive integers.

[0131] According to one embodiment of the present invention, This is the ratio of the historical weather severity coefficient at time j in the i-th historical control cycle to the preset historical weather severity coefficient threshold. The larger this ratio, the larger the historical weather severity coefficient. The preset historical weather severity coefficient threshold can be set to 0.5. This is the ratio of the historical queue size at time j in the i-th historical control cycle to the preset historical queue size threshold. The larger this ratio, the larger the historical queue size. The preset historical queue size threshold can be set to 10 people. This is the ratio of the number of participants in a historical building event at time j of the i-th historical control period to the preset threshold for the number of participants in a historical building event. The larger this ratio, the more participants there are in the historical building event. The preset threshold for the number of participants in a historical building event can be set to 10. This is the ratio of the historical proportion of operable elevators at time j in the i-th historical control cycle to the preset operable elevator proportion threshold. The larger this ratio, the larger the historical proportion of operable elevators. The preset operable elevator proportion threshold can be set to 1. For the historical working day identification result of the i-th historical control period, if the date of the i-th historical control period is not a working day, then... If the date of the i-th historical control cycle is a working day, then... This is the ratio of the average historical passenger flow of the elevator over multiple weekdays to the average historical passenger flow of the elevator over multiple weekends and holidays, and this ratio must be greater than 1. The larger the elevator, the greater the passenger flow on weekdays compared to weekends and holidays. The historical operational readiness identification result of the k-th elevator at the j-th moment of the i-th historical control cycle is given. When the k-th elevator is operational... The value is 1, when the k-th elevator is not running. =0, The historical weather severity coefficient, historical queue length, historical building event attendance, historical weekday identification results, and historical elevator availability identification results are positively correlated with the predicted sample passenger flow. Conversely, the historical elevator availability ratio is negatively correlated with the predicted sample passenger flow. For example, a higher historical weather severity coefficient indicates more severe weather conditions; rain, high temperatures, extreme cold, and strong winds encourage people to stay indoors, reducing their time outdoors and concentrating passenger flow at building entrances, thus increasing the predicted sample passenger flow. Similarly, a larger historical queue length results in a larger queue at elevator entrances and more people using elevators, leading to a larger predicted sample passenger flow. Furthermore, a larger historical building event attendance indicates more people will use elevators after meetings or events, further increasing the predicted sample passenger flow. Finally, weekday passenger flow is generally higher than that on weekends and holidays. The larger the value, the larger the predicted sample passenger flow. For example, when When the value is 1, the elevator can operate normally. When the value is 0, the elevator cannot operate normally. The passenger flow when the elevator is not operating normally is 0. The passenger flow when the elevator is operating normally is greater than the passenger flow when it is not operating normally. The larger the value, the larger the predicted sample passenger flow. For example, when the proportion of historically operational elevators is higher, there are more elevators that can operate normally, but each elevator handles less passenger flow, resulting in a lower predicted sample passenger flow. Therefore, , , , and The larger the value, the larger the predicted sample passenger flow. The larger the value, the smaller the predicted sample passenger flow.

[0132] According to one embodiment of the present invention, Let the relative error between the predicted sample passenger flow and the historical passenger flow of the k-th elevator at time j in the i-th historical control cycle be used. A weighted average is used to calculate the relative error between the predicted sample passenger flow and the historical passenger flow, resulting in a training loss function. During training, this training loss function is reduced, thereby decreasing the relative error between the predicted sample passenger flow and the historical passenger flow, improving the accuracy of the passenger flow prediction model, and ultimately enhancing the overall accuracy of the passenger flow prediction model.

[0133] In this way, the training loss function of the passenger flow prediction model can be determined based on historical passenger flow, predicted sample passenger flow, historical proportion of operational elevators, historical operational identification results, historical weekday identification results, historical weather severity coefficient, number of participants in historical building events, and historical queuing length. During the calculation process, the possible impact of the proportion of operational elevators, weekday identification results, weather severity coefficient, number of participants in building events, and queuing length on the predicted passenger flow can be determined to assess the error of the predicted sample passenger flow. Based on this impact and the relative error of the predicted sample passenger flow, the training loss function can be set to reduce the training loss function during the training process, thereby more effectively improving the accuracy of the passenger flow prediction model.

[0134] According to one embodiment of the present invention, in step S8, the power allocation coefficient is determined based on the predicted passenger flow.

[0135] For example, when the predicted passenger flow is greater than or equal to the preset passenger flow threshold (which can be set to 100 people), it means that the elevator is predicted to be busy in the next five minutes, and the power allocation coefficient is 1. Conversely, it means that the elevator is predicted to be relatively idle in the next five minutes, and the power allocation coefficient is 0.

[0136] According to one embodiment of the present invention, in step S9, power dispatch control is performed based on the power allocation coefficient.

[0137] For example, when the elevator's power allocation coefficient is 1, it indicates that the elevator is predicted to be busy within the next five minutes. Within the total safety limit, the maximum pre-allocated power is allocated to the elevator. For instance, if the elevator's peak demand is 50kW, then 50kW of power is pre-allocated. When the elevator's power allocation coefficient is 0, no pre-allocated power is needed. Pre-allocated power allows the elevator motor to accelerate faster, directly shortening the time of a single trip. This allows for the transport of more passengers during peak hours, significantly reducing waiting times and making power allocation more planned and predictable. It also reduces the competition for peak power resources among multiple elevators, minimizing the impact on the power grid and ensuring more stable system operation.

[0138] The energy storage and retrieval control method based on big data according to embodiments of the present invention can accurately analyze whether energy recovery is possible under the current elevator and power grid conditions based on elevator control signals, elevator operating status information, and power grid status information. It determines energy recovery discrimination parameters, stores the recovered energy when recovery is possible, and predicts passenger flow for the elevator in the next five minutes using a passenger flow prediction model. Based on the predicted passenger flow, it determines whether to retrieval control of the recovered energy, thus improving the effectiveness of energy storage and retrieval control. When determining historical weather severity coefficients, historical weather data can be used. During the calculation process, weather severity can be assessed based on temperature, wind speed, and rainfall, improving the comprehensiveness and accuracy of historical weather severity coefficients. When determining the training loss function for the passenger flow prediction model, it can be based on historical passenger flow, predicted sample passenger flow, historical proportion of operational elevators, historical operational identification results, historical weekday identification results, historical weather severity coefficient, number of participants in historical building events, and historical queuing numbers. During the calculation process, the potential impact of the proportion of operational elevators, weekday identification results, weather severity coefficient, number of participants in building events, and queuing numbers on the predicted passenger flow can be considered to determine the influence of the above data on the error of the predicted sample passenger flow. Based on this influence and the relative error of the predicted sample passenger flow, the training loss function can be set to reduce the training loss function during the training process, thereby more effectively improving the accuracy of the passenger flow prediction model.

[0139] Figure 3 illustrates, exemplarily, a block diagram of a big data-based power storage and retrieval control system according to an embodiment of the present invention, the system comprising:

[0140] The integrated information module is used to acquire elevator cabinet control signals and power grid status information during the control cycle;

[0141] The operation information module is used to acquire elevator operation status information during the control cycle;

[0142] The discrimination parameter module is used to determine the energy recovery discrimination parameters based on the elevator operating status information, the elevator cabinet control signal, and the power grid status information.

[0143] The energy recovery discrimination module is used to determine whether to perform energy recovery based on the energy recovery discrimination parameters.

[0144] An energy recovery module is used to store recovered energy when it is necessary to recover energy.

[0145] The real-time data module is used to acquire real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images;

[0146] The passenger flow prediction module is used to process the elevator operation status information, the real-time date data, the real-time weather data, the building event scheduling data, and the real-time monitoring images based on the trained passenger flow prediction model to determine the predicted passenger flow.

[0147] The allocation coefficient module is used to determine the power allocation coefficient based on the predicted passenger flow.

[0148] The control module is invoked to control the energy allocation based on the energy distribution coefficient.

[0149] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0150] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are merely examples and do not limit the present invention. The objectives of the present invention have been fully and effectively achieved. The functions and structural principles of the present invention have been demonstrated and explained in the embodiments, and any variations or modifications may be made to the implementation of the present invention without departing from the stated principles.

Claims

1. A method for energy storage and retrieval control based on big data, characterized in that, include: During the control cycle, the elevator cabinet control signal and power grid status information are acquired; during the control cycle, the elevator operating status information is acquired; based on the elevator operating status information, the elevator cabinet control signal, and the power grid status information, energy recovery discrimination parameters are determined; based on the energy recovery discrimination parameters, it is determined whether to perform energy recovery; if energy recovery is required, the recovered energy is stored. Acquire real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images; process the elevator operation status information, real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images according to a trained passenger flow prediction model to determine the predicted passenger flow; determine the power allocation coefficient based on the predicted passenger flow. Based on the aforementioned power allocation coefficient, power dispatch control is performed; Based on the elevator operating status information, the elevator cabinet control signal, and the power grid status information, energy recovery discrimination parameters are determined, including: determining a first recovery discrimination result based on the elevator cabinet control signal; determining the DC bus voltage, grid current harmonics, power factor, grid voltage fluctuation, frequency deviation, and battery capacity based on the power grid status information; determining a second recovery discrimination result based on the DC bus voltage, grid current harmonics, power factor, grid voltage fluctuation, frequency deviation, and battery capacity; determining the elevator mass, elevator load, and elevator descent height based on the elevator operating status information; determining a third recovery discrimination result based on the elevator mass, elevator load, and elevator descent height; determining energy recovery discrimination parameters based on the first recovery discrimination result, the second recovery discrimination result, and the third recovery discrimination result; and determining the energy recovery discrimination parameters based on the DC bus voltage, the grid current harmonics, and the power factor. The process involves determining a second recovery discrimination result based on the power factor, the grid voltage fluctuation, the frequency deviation, and the battery capacity. This includes: determining a voltage and current constraint result based on the DC bus voltage, the grid current harmonics, and the power factor; determining a grid state constraint result based on the grid voltage fluctuation and the frequency deviation; determining a battery state constraint result based on the battery capacity; and determining the second recovery discrimination result based on the voltage and current constraint result, the grid state constraint result, and the battery state constraint result. A third recovery discrimination result is then determined based on the elevator mass, the elevator load, and the elevator descent height. This includes: determining the expected recovery power based on the elevator mass, the elevator load, and the elevator descent height; obtaining the rated power of the elevator motor; determining a preset recovery power threshold; and determining the third recovery discrimination result based on the expected recovery power, the rated power of the elevator motor, and the preset recovery power threshold.

2. The method for energy storage and retrieval control based on big data according to claim 1, characterized in that, The training steps of the passenger flow prediction model include: acquiring historical passenger flow, historical elevator operation status information, historical date data, historical weather data, historical building event scheduling data, and historical monitoring images for historical control cycles; determining the historical proportion of operable elevators and historical operability identification results based on the historical elevator operation status information; determining the historical weekday identification results based on the historical date data and historical passenger flow; determining the historical weather severity coefficient based on the historical weather data, wherein the historical weather data includes: historical temperature data, historical wind speed data, and historical rainfall data; determining the number of participants in historical building events corresponding to multiple moments in the historical control cycle based on the historical building event scheduling data; and determining the number of participants in historical building events based on the historical monitoring images. Determine the historical queue length; process the historical operational availability identification results, the historical operational elevator ratio, the historical weekday identification results, the historical weather severity coefficient, the historical building event participants, and the historical queue length based on the passenger flow prediction model to obtain the predicted sample passenger flow; determine the training loss function of the passenger flow prediction model based on the historical passenger flow, the predicted sample passenger flow, the historical operational elevator ratio, the historical operational availability identification results, the historical weekday identification results, the historical weather severity coefficient, the historical building event participants, and the historical queue length; train the passenger flow prediction model based on the training loss function to obtain the trained passenger flow prediction model.

3. The method for energy storage and retrieval control based on big data according to claim 2, characterized in that, Based on the historical weather data, the historical weather severity coefficient is determined, including: according to the formula: Determine the historical weather severity coefficient at time j in the i-th historical control cycle. ,in, This provides the historical temperature data for the j-th moment of the i-th historical control cycle. To preset historical temperature data thresholds, This provides the historical wind speed data for the j-th moment of the i-th historical control cycle. To preset historical wind speed data thresholds, For the historical rainfall data at time j in the i-th historical control cycle, This is a preset threshold for historical rainfall data.

4. The energy storage and dispatch control method based on big data according to claim 2, characterized in that, Based on the historical passenger flow, the predicted sample passenger flow, the historical proportion of operational elevators, the historical operational elevator identification results, the historical weekday identification results, the historical weather severity coefficient, the number of participants in historical building events, and the historical queuing length, the training loss function of the passenger flow prediction model is determined, including: according to the formula: Determine the training loss function for the passenger flow prediction model. ,in, Let the predicted sample passenger flow of the k-th elevator be the passenger flow at the j-th time point in the i-th historical control cycle. Let the historical passenger flow of elevator k be the passenger flow from time j in the i-th historical control cycle to five minutes after time j. This represents the proportion of historically operable elevators at time j in the i-th historical control cycle. To preset the threshold for the proportion of elevators that can operate, This represents the historical operational availability identification result of the k-th elevator at the j-th moment of the i-th historical control cycle. ， Let be the historical weather severity coefficient at time j in the i-th historical control cycle. To preset the historical weather severity coefficient threshold, Let be the historical queue size at time j in the i-th historical control cycle. To preset a historical threshold for the number of people in the queue, The number of participants in the historical building event at time j of the i-th historical control period. To preset a threshold for the number of participants in historical building events, The result of identifying the historical working day for the i-th historical control cycle is given, where n is the number of historical control cycles (i ≤ n), j is the number of moments in the historical control cycle (j ≤ m), and i, n, j, and m are all positive integers.

5. A power storage and retrieval control system based on big data, characterized in that, The method for controlling energy storage and retrieval based on big data, as described in any one of claims 1-4, is operated as follows: the control system for energy storage and retrieval based on big data comprises: a comprehensive information module for acquiring elevator cabinet control signals and power grid status information during the control cycle; an operation information module for acquiring elevator operation status information during the control cycle; a discrimination parameter module for determining energy recovery discrimination parameters based on the elevator operation status information, the elevator cabinet control signals, and the power grid status information; a energy recovery discrimination module for determining whether to perform energy recovery based on the energy recovery discrimination parameters; and an energy recovery module... The system comprises the following modules: a storage module for storing recovered energy when energy recovery is required; a real-time data module for acquiring real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images; a traffic prediction module for processing the elevator operating status information, real-time date data, real-time weather data, building event scheduling data, and real-time monitoring images based on a trained passenger flow prediction model to determine the predicted passenger flow; an allocation coefficient module for determining the energy allocation coefficient based on the predicted passenger flow; and a call control module for performing energy call control based on the energy allocation coefficient.

Images