A device for obtaining the number of outbound passengers in a vertical passenger elevator

Abstract

The present invention relates to the technical field of elevator group control, and discloses a device for obtaining the number of outbound passengers of a vertical passenger elevator, comprising a passenger number input module, a passenger number prediction module and a passenger number integration module; the passenger number input module detects the pressing time of a pressure-sensitive button by a passenger to obtain the number of outbound passengers input by the elevator; the passenger number prediction module obtains the predicted number of outbound passengers according to the running state of the elevator; the passenger number integration module integrates and calculates the passenger input number and the passenger prediction number to obtain the number of outbound passengers of the elevator; the present invention obtains the number of outbound passengers by integrating and calculating the number of passengers input and the predicted number of passengers; M1 The present invention can obtain the number of passengers waiting for the elevator on the current floor in real time, providing a reliable basis for the instant optimization of the elevator group control strategy. In addition, the high-confidence data provided by the passenger number input module can be used as a passenger number prediction module. M2 The supervised learning labels promote the continuous optimization of the associated model and improve the prediction accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of elevator group control, and in particular to a device for obtaining the number of outbound passengers in a vertical passenger elevator. Background Art

[0002] In modern buildings, the optimization of elevator group control systems is crucial to improving elevator operation efficiency and passenger satisfaction. Traditional elevator dispatching systems usually rely on preset fixed rules (such as the publication number CN115973866A, the patent name is: an elevator group control dispatching system; the publication number CN116620977A, the patent name is: a six-unit ten-story elevator pick-up method based on fuzzy control), but these rules are not flexible enough when facing the actual passenger demand during peak hours, and it is difficult to effectively adapt to changes. In recent years, the application of destination selectors has begun to change the elevator group control strategy. Compared with the traditional up / down button elevator call mode, the destination selector elevator call mode can accurately obtain parameters such as the number of waiting passengers at the time of elevator call on the floor where the selector is installed and the destination floor of the waiting passengers. Based on this, some studies have tried to predict passenger flow based on historical data and adjust the elevator dispatching strategy (such as the publication number CN116605731A, the patent name is: a hybrid elevator group control system based on prediction estimation), which significantly improved the efficiency of elevator group control. However, the cost of this efficiency improvement is to install a destination selector on each floor, which significantly increases the cost. In order to control costs, most buildings only install a destination selector on the first floor, and use the traditional up / down button call method on other floors. Therefore, how to accurately obtain the number of passengers called out by the vertical passenger elevator has become the key to optimizing the elevator group control system. Some studies have attempted to obtain the number of passengers through machine vision (such as the publication number: CN113697620A, the patent name is a method and device for elevator group control based on visual recognition; the publication number: CN117902415A, the patent name is: a method and device for intelligent elevator monitoring based on the Internet of Things; the publication number: CN116924167A, the patent name is: a method and device for improving elevator transportation efficiency based on the number of people inside and outside the elevator), but practical applications are often limited by privacy issues and the complexity of data processing. In addition, some progress has been made in the use of machine learning technology to predict the number of elevator passengers (such as the publication number: CN117985557A, the patent name: A method and system for controlling elevator clusters). Such methods usually require large-scale data training for specific application scenarios, resulting in limited universality and accuracy of the model. Their adaptability to emergencies or non-standard situations also needs to be improved. Therefore, there is a need to propose a more accurate and cost-effective method for obtaining the number of outbound passengers that is applicable to both destination floor registration and up and down registration elevator call modes. Summary of the invention

[0003] The problem to be solved by the present invention is how to effectively obtain the number of outbound passengers in a vertical passenger elevator. Based on this, the present invention proposes a device for obtaining the number of outbound passengers in a vertical passenger elevator. The device includes a passenger number input module M1 , Passenger Number Prediction Module M2 Integration module with passenger count M3 .

[0004] Furthermore, the passenger number input module M1 Detect the length of time a passenger presses a pressure-sensitive button to get the number of passenger inputs for the elevator call N I . Passenger Number Prediction Module M2 Get the predicted number of elevator outbound passengers based on the elevator operation status N p . Passenger count integration module M3 Integrate the number of passengers input and the number of passengers predicted to get the number of passengers called out by the elevator N .

[0005] Furthermore, the passenger number input module M1 Including pressure sensing unit S22 and passenger conversion unit S21 . Pressure Sensing Unit S22 Collect the passenger's pressing signal, mainly including the pressing duration, and calculate the confidence of the passenger input number. Passenger number conversion unit S21 Convert the pressing duration of this pressing action into the number of outbound passenger inputs N I . Number of outbound passenger inputs N I Passed to the passenger count integration module.

[0006] Furthermore, the passenger number prediction module M2 It includes a data collection unit and a passenger number prediction unit. The data collection unit obtains the elevator operation parameters such as the door opening time, door closing time, and operation floor in the previous time period. The passenger number prediction unit calculates the predicted number of outbound passengers based on the association model. N p . Forecast number of outbound passengers N p Passed to the passenger count integration module M3 .

[0007] Furthermore, the association model building step includes: 1) converting the elevator operation parameters of the data acquisition unit into R p as input; 2) Enter the number of passengers into the module M1 The number of outbound passenger inputs with confidence greater than the threshold N Ias output label; 3) construct input parameters R p Passenger input number N I The association model uses gradient descent or maximum likelihood algorithm to obtain the model input parameters R p Passenger input number N I The best fit of .

[0008] Furthermore, when a passenger uses a pressure-sensitive button to call the elevator, the passenger number integration module M3 The operation performed is to convert the predicted number of passengers N p Passenger input number N I Perform weighted calculation to get the number of passengers called out by the elevator N The weighted calculation step includes: 1) inputting the number of passengers into the module M1 Get the number of passengers input by the elevator; 2) From the passenger number prediction module M2 Get the predicted number of elevator outbound passengers; 3) According to the formula Calculate the number of passengers called out by the elevator N， Among them, λ is a weight factor dynamically determined according to the pressing time input by the passenger. The longer the pressure-sensitive button is pressed, the more outbound passengers the user has input, and the more reliable the prediction is.

[0009] The present invention also provides an elevator group control strategy optimization system based on a vertical passenger elevator outbound call passenger number acquisition device, comprising a car state calculation module, a call allocation module, a car control module, and a call data and car state historical data storage module;

[0010] The car status calculation module is responsible for integrating the number of passengers from the M3 Number of passengers receiving elevator calls N , and combined with the elevator control signal provided by the car control module to determine the current state of the elevator car, and transmit the relevant data to the call distribution module to ensure the coordination and efficiency of the elevator operation;

[0011] The call distribution module is responsible for coordinating and controlling the entire elevator group control system, integrating the information from the car status calculation module and the passenger quantity data, generating elevator dispatch instructions, and ensuring that the instructions are correctly executed;

[0012] The car control module performs specific car dispatching operations according to the dispatching strategy of the call distribution module. After execution, the module feeds back the operation results to the car status calculation module in real time to continuously optimize the response and dispatching strategy of the elevator.

[0013] The call data and car status historical data storage module is responsible for data archiving and subsequent processing, recording the situation of each elevator dispatch and the output data of the passenger number integration module; the recorded data will be used to train and optimize the algorithm of the passenger prediction unit in the passenger number prediction module, continuously improving its prediction accuracy and the overall performance of the system.

[0014] The invention is beneficial in that: by inputting the number of passengers into the module M1 The present invention can obtain the number of passengers waiting for the elevator on the current floor in real time, providing a reliable basis for the instant optimization of the elevator group control strategy. Especially during peak hours or emergencies, it can accurately judge the current passenger demand and elevator status and achieve rapid response. In addition, the passenger number input module M1 The high confidence data provided can be used as a passenger number prediction module M2 The supervised learning labels of the association model promote the continuous optimization of the association model. This self-learning ability allows the association model to continuously improve its accuracy during use without the need to collect a large amount of training data in advance, and is applicable to buildings of different types and sizes. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a reference diagram of the design idea of ​​the present invention;

[0016] Figure 2 It is a system structure diagram of the present invention and a pressure sensitive button device structure diagram;

[0017] Figure 3 It is a structural diagram of the elevator group control strategy optimization system of the present invention;

[0018] Figure 4 This is a flow chart of the optimization process of predicting the number of outbound passengers of the elevator according to the present invention;

[0019] Figure 5 Describe schematic diagrams for specific application scenarios;

[0020] Figure 6 This is a visualization sample diagram of the output results of the passenger quantity prediction model of the present invention. DETAILED DESCRIPTION

[0021] With reference to the accompanying drawings, the present invention relates to a device for obtaining the number of outbound passengers in a vertical passenger elevator. The following is a description of a specific embodiment of the present invention.

[0022] The present invention proposes a device for obtaining the number of outbound passengers in a vertical passenger elevator, aiming at effectively obtaining the number of outbound passengers in a vertical passenger elevator. Figure 1 The invention is a description of the improvements that can be achieved by adopting a device for obtaining the number of outbound passengers in a vertical passenger elevator designed by the present invention.

[0023] Take the rush hour of an 8-story office building as an example. The building's elevator system is equipped with two elevators, each with a rated passenger capacity of 15 people. At this time, there are 10 people on the 8th floor ready to go down to the 1st floor, and 10 people on the 6th floor also ready to go down to the 1st floor. The situation of the two cars is that car A is currently on the 8th floor and has picked up 10 passengers. Car B is idle.

[0024] When the up and down buttons are used to call the elevator, the passenger on the 6th floor calls the elevator by pressing the down button. Since the elevator call distribution module cannot know the specific number of people on this floor, car A responds to the down command of the passenger on the 8th floor and goes down to the 6th floor. At this time, car A can only take 5 more people because it already carries 10 people, so the remaining 5 people cannot get on the car. These passengers need to press the down button again to wait for the elevator. Car B responds to the new down command and goes up from the 1st floor to the 6th floor.

[0025] When a vertical passenger elevator outbound passenger number acquisition device designed by the present invention is used, the passenger on the 8th floor indirectly inputs the number of passengers at this time as 10 by pressing the device of the present invention for a long time. The elevator call distribution module directly dispatches the idle car B from the 1st floor to the 6th floor to respond to the demand based on the detailed passenger information input in real time. The present invention enables the elevator group control system to optimize the scheduling strategy more reasonably by accurately acquiring the number of outbound passengers.

[0026] Figure 2 It is a system structure diagram of the present invention and a pressure sensitive button device structure diagram.

[0027] According to its characteristics, the device includes a passenger number input module M1 , Passenger Number Prediction Module M2 Integration module with passenger count M3 .

[0028] Passenger Number Input Module M1 Responsible for detecting the length of time a passenger presses the elevator call button to obtain the number of outbound passenger inputs for a certain call N I The passenger number input module includes a pressure sensing unit S22 and passenger conversion unit S21 . Pressure Sensing Unit S22 Collect the passenger's pressing signal, mainly including the pressing duration, and calculate the confidence of the passenger input number. Passenger number conversion unit S21 According to the pressing time of this pressing action, the corresponding pressing time is converted into the outbound passenger input number. N I . Number of outbound passenger inputs N I Passed to the passenger count integration module.

[0029] Passenger Number Prediction Module M2 According to the historical operation data of the elevator door machine obtained from the control system mainboard, the predicted number of passengers in the elevator during the current outbound call is predicted. N p The historical operation data of the elevator door machine includes but is not limited to the door opening time, door closing time and arrival floor. This module includes a data acquisition unit and a passenger number prediction unit. The data acquisition unit obtains the elevator operation parameters such as the door opening time, door closing time, and operation floor in the previous time period. The passenger number prediction unit calculates the predicted number of passengers according to the association model. N p . Passenger forecast N p Passed to the passenger count integration module.

[0030] Passenger count integration module M3 The predicted number of passengers N p Passenger input number N I Integrate to get the final predicted number of elevator outbound passengers N When a passenger uses a pressure-sensitive button to call an elevator, the passenger count integration module first evaluates the length of time the passenger presses the button to determine the confidence level of the passenger input number. By weighted integration of the predicted number of passengers and the input number of passengers, this module can accurately calculate the expected number of passengers for the elevator to call. N This result will be used to adjust the distribution strategy.

[0031] The passenger number input module is the core device of the present invention. The physical structure of the device consists of four core components: mounting bracket S1 , Pressure detection processing unit S2 , User Interface Buttons S3 (pressure-sensitive buttons), and a halo display indicator S4 .

[0032] The mounting bracket is firmly mounted on the elevator call panel of each floor through four threaded holes and is equipped with eight external threaded studs to ensure a stable connection with the pressure detection processing unit. This structural design ensures the safety and stability of the equipment in daily use.

[0033] Pressure detection processing unit includes passenger number conversion unit S21 and pressure detection unit S22 The pressure detection unit is responsible for sensing the pressure of the up and down buttons and converting the pressure into electrical signals. The passenger number conversion unit receives these electrical signals and calculates the estimated number of passengers based on the pressing duration, acting as the data processing core of the intelligent input module.

[0034] The user interface buttons include an up / down selection function, allowing the user to estimate the number of passengers by how long the button is pressed. These buttons serve as a direct interface for the user to interact with the system and are easy to operate.

[0035] The halo display indicator is responsible for providing instant visual feedback when the user operates. The indicator indicates the number of passengers added by the number of halo rings that light up one by one through the pressing action, and each complete bright ring represents the addition of one passenger. This visual feedback mechanism not only enhances the user experience, but also intuitively displays the number of passengers entered, thereby improving the interactivity and accuracy of the system.

[0036] This device effectively captures and processes information input by users, provides key data for elevator group control strategies, and greatly optimizes the dispatching efficiency and accuracy of elevators.

[0037] Figure 3 The structure diagram of the elevator group control strategy optimization system in the hybrid elevator call mode of the present invention is shown. The whole system includes a passenger number prediction module, a passenger number input module, a passenger number integration module, a car state calculation module, a call distribution module, a car control module and a call data and car state history data storage module.

[0038] The passenger number input module detects the length of time the passenger presses the pressure-sensitive button to obtain the number of passengers input for the elevator call. N I The passenger number prediction module obtains the predicted number of elevator outbound passengers based on the elevator operation status. N p The passenger number integration module integrates the passenger input number and the passenger prediction number to obtain the number of elevator outbound passengers. N .

[0039] The car status calculation module is responsible for receiving the number of elevator outbound passengers from the passenger number integration module and determining the current state of the elevator car in combination with the elevator control signal provided by the car control module. This module transmits the relevant data to the call allocation module to ensure the coordination and efficiency of the elevator operation.

[0040] The call distribution module is the command center of the system and is responsible for coordinating and controlling the entire elevator group control system. This module integrates the information from the car status calculation module and the passenger quantity data to generate elevator dispatch instructions and ensure that these instructions are executed correctly.

[0041] The car control module performs specific car dispatch operations according to the call allocation module's dispatch strategy. After execution, the module feeds back the operation results to the car status calculation module in real time to continuously optimize the elevator's response and dispatch strategy.

[0042] The call data and car status historical data storage module plays the role of data archiving and subsequent processing, recording the details of each elevator dispatch and the output data of the passenger number integration module. The recorded data will be used to train and optimize the algorithm of the passenger prediction unit in the passenger number prediction module, continuously improving its prediction accuracy and the overall performance of the system.

[0043] Figure 4 It is a flow chart of the elevator group control strategy optimization process of the present invention.

[0044] At present, the elevator group control system mainly uses two basic elevator calling methods: up / down button calling and destination selector calling. If the passenger chooses to use the pressure-sensitive button to call the elevator, the system will estimate the number of passengers calling the elevator by analyzing the pressing time. If the passenger uses the destination selector to call the elevator, he needs to directly enter the destination floor and the number of accompanying people. The number of outbound passengers determined by the above two calling methods will be used as the input passenger number for the prediction of the number of outbound passengers.

[0045] At the same time, whenever there is a new elevator call, the passenger number prediction module estimates the number of outbound passengers for this elevator call based on historical operation data. This prediction involves multiple parameters, such as the elevator door opening and closing time and the operating floor. This process can be described by formula 1:

[0046]

[0047] in, To predict the number of passengers, The elevator operation parameters are door opening time, door closing time, and action floor. It is a variety of neural network models suitable for regression tasks.

[0048] The predicted number of passengers obtained by the passenger number prediction module and the number of input passengers directly obtained through the two elevator call methods will be combined according to the following formula to form the final passenger prediction value N , the formula is as follows:

[0049]

[0050] here, is the predicted number of passengers based on the elevator operation status prediction, The number of passenger inputs directly obtained based on the two elevator call methods. Weight coefficient It can be dynamically adjusted according to the pressing time of the pressure-sensitive button to reflect the confidence of the number of passengers directly input by the two elevator calling methods. The longer the button is pressed, the more passengers the user inputs, which also means that the prediction has a higher reliability. The specific value can be dynamically adjusted through the following formula:

[0051]

[0052] in T Indicates the duration of pressing a pressure-sensitive button in seconds. To set the press duration threshold.

[0053] The final predicted number of passengers N The input is sent to the car status calculation module, which is responsible for optimizing the dispatch of the car according to the car status and the predicted number of passengers, such as controlling the car to respond to calls based on the minimum average waiting time or the lowest energy consumption ratio.

[0054] In addition, after each elevator call response is completed, the passenger prediction unit in the passenger number prediction module will record the data of this run and calculate the deviation between the predicted number of passengers and the input number of passengers determined by the two elevator call methods. This information is back-propagated to the model and trained by using optimizers such as Adam or SGD to further improve the accuracy of the prediction. The loss function used in training is usually Mean-Square Error (MSE) or Root-Mean-Squared Error (RMSE), calculated as:

[0055]

[0056] in, is the training step length, is the predicted number of passengers based on the elevator operation status prediction, It is the number of passenger inputs directly obtained based on the two elevator call methods. By minimizing the loss, the model can effectively learn the complex relationship between the elevator operation status and the number of outbound passengers. This process not only improves the accuracy of the prediction, but also enables the model to adapt to the ever-changing elevator usage pattern.

[0057] The deployment of the present invention will provide a more dynamic and adaptive solution for elevator operation management, enabling the elevator system to better cope with changes in passenger flow during peak hours while maintaining efficient operation, ensuring a safe and convenient passenger experience.

[0058] The following is a specific description of the application of the present invention in a commercial building with multiple elevators. Figure 5 A schematic diagram is described for this application scenario.

[0059] During the evening rush hour in a 12-story commercial building, the elevator system often faces a large number of downlink requests. The building's elevator system is equipped with three elevators, A, B, and C, each with a rated passenger capacity of 15 people. Initially, all elevators are idle. Elevator A is on the 1st floor, elevator B is on the 5th floor, and elevator C is on the 8th floor. At this time, there are 20 people on the 12th floor ready to go down, and 15 people on the 10th floor are also ready to go down.

[0060] 20 people on the 12th floor use the pressure-sensitive button to call the elevator. The pressure-sensing unit of the passenger number input module detects that the user has been pressing for 6 seconds. According to the system settings, 2 people are estimated to be pressed per second, so the signal conversion unit estimates that there are 12 people on the 12th floor who need service. 15 people on the 10th floor also use the pressure-sensitive button to call the elevator. They press it for 5 seconds, and it is estimated that 10 people need service. The above data are all transmitted to the passenger number prediction module through the data output unit.

[0061] At the same time, the elevator passenger number prediction module receives the historical operation data within a week, including the elevator door opening and closing time and its operation floor. The model input data is shown in Table 1.

[0062] Table 1 Passenger quantity prediction model training data input sample

[0063]

[0064] At this time, the model uses a multi-layer perceptron. The passenger prediction unit of the passenger number prediction module predicts that the predicted passenger numbers on the 12th and 10th floors are 12 and 10 respectively, and transmits the predicted data to the passenger number integration module through the data output unit.

[0065] The weight adjustment unit in the passenger number integration module receives the pressing durations of the two calls from the passenger number input module as 6S and 5S respectively. The system sets the corresponding relationship between the pressing duration and the weight system as

[0066]

[0067] According to the setting, the weight adjustment unit determines the weight coefficient of the two calls. Both are 1. Finally, the passenger number integration module calculates and predicts that there are 12 people on the 12th floor who need to be served and 10 people on the 10th floor who need to be served. Since elevator C is closest to the 12th floor, it is dispatched to the 12th floor to pick up passengers. Elevator B, as the elevator closest to the 10th floor, is dispatched to respond to the needs of the 10th floor, while elevator A remains on standby on the 1st floor in case of emergency.

[0068] After each elevator call, the prediction module records the operation data and the actual number of passengers, and calculates the predicted deviation. These data will be used for further training of the model. Here, MSE is used as the loss function, and Adam is selected as the optimizer to fine-tune the model parameters and improve the accuracy of the prediction.

[0069] The model is continuously trained, and the comparison between the output of the passenger number prediction model and the input passenger number is shown in Table 2;

[0070] Table 2 Comparison example between the output results of the passenger number prediction model and the input passenger number

[0071]

[0072] Figure 6 Sample visualization of the output of the passenger number prediction model.

[0073] The speed of response of the elevator dispatch system in emergency situations is critical when faced with changes in passenger flow. The pressure sensitive button device of the present invention provides an innovative solution to enhance the efficiency and responsiveness of the elevator system in emergency situations.

[0074] The following is a specific application description of the present invention in an emergency situation in a 15-story hotel.

[0075] In a 15-story hotel equipped with two elevators (A and B) with a rated passenger capacity of 15 people, the elevator group control system adopts a hybrid elevator calling method, that is, only the destination selector is installed on the first floor, and the traditional up / down button elevator calling method is used on other floors. At this time, elevator A is carrying 5 passengers from the 10th floor to perform a down mission, while elevator B is on standby on the 1st floor, ready to receive new instructions.

[0076] As a large conference was over, a large number of participants quickly gathered on the hotel's conference room floor (7th floor) and prepared to go downstairs. Since there was no destination selector installed on this floor, guests on the 7th floor could only use the standard down button to call the elevator. The elevator dispatch system might default to calling the nearest elevator A that was performing a down mission to respond to this call, while elevator B remained on standby.

[0077] However, using the pressure-sensitive button of the present invention, the participant conveys an urgent and high-priority elevator call by pressing the button for a long time. The system captures the long-term press through the pressure sensing unit of the device and quickly determines it as a high-urgency call, and predicts that the number of passengers in this call far exceeds the remaining capacity of elevator A.

[0078] Therefore, the system makes a quick dispatch decision and directly dispatches the idle elevator B from the 1st floor to the 7th floor, ensuring that the guests can be taken to their destination as quickly as possible. This decision fully demonstrates the flexibility and rapid response capabilities of the present invention, and significantly improves the timeliness and accuracy of elevator services by efficiently utilizing the data analysis function of the pressure-sensitive button.

[0079] The contents described in the embodiments of this specification are merely an enumeration of the implementation forms of the inventive concept. The protection scope of the present invention should not be regarded as limited to the specific forms described in the embodiments. The protection scope of the present invention also extends to equivalent technical means that can be conceived by those skilled in the art based on the inventive concept.

Claims

1. A device for obtaining the number of outbound passengers in a vertical passenger elevator, characterized in that: Includes passenger count input module M1 , Passenger Number Prediction Module M2 Integration module with passenger count M3 ; The passenger number input module M1 Responsible for detecting the length of time that passengers press the elevator call button to obtain the number of passengers waiting for the elevator at the current time. N I ; Passenger number input module M1 The number of outbound passenger inputs obtained N I Passenger number prediction module M2 The model provides supervision labels to implement the passenger number prediction module M2 Continuous learning; The passenger number prediction module M2 According to the historical operation data of the elevator door machine obtained from the control system mainboard, the predicted number of outbound passengers waiting for the elevator at the current time is predicted. N p ; Elevator door machine historical operation data includes door opening time, door closing time and arrival floor; The passenger number integration module M3 Enter the number of outbound passengers N I Forecast number of outbound passengers N p Integration to get the number of passengers called out by the elevator N .

2. A device for obtaining the number of outbound passengers in a vertical passenger elevator according to claim 1, characterized in that The passenger number input module M1 Including pressure sensing unit S22 and Passenger Conversion Unit S21 ; Pressure sensing unit S22 Collect the pressing signal of the passenger, including the pressing duration. The longer the pressing duration, the more passengers the user inputs, so as to obtain the confidence of the number of passengers input; Passenger number conversion unit S21 Convert the pressing time into the number of outbound passenger inputs N I , and input the number of outbound passengers N I Passed to the passenger count integration module.

3. A device for obtaining the number of outbound passengers in a vertical passenger elevator according to claim 1, characterized in that The passenger number prediction module M2 It includes a data acquisition unit and a passenger number prediction unit; the data acquisition unit obtains the elevator operation parameters in the previous time period R p , including door opening time, door closing time, and action floor; the passenger number prediction unit calculates the predicted number of outbound passengers based on the association model N p , and the predicted number of outbound passengers N p Passed to the passenger count integration module M3 .

4. A device for obtaining the number of outbound passengers in a vertical passenger elevator according to claim 3, characterized in that The association model building step includes: 1) The elevator operation parameters of the data acquisition unit R p As input; 2) Input the number of passengers into the module M1 The number of outbound passenger inputs with confidence greater than the threshold N I As output label; 3) Construct input parameters R p Number of outbound passengers input N I The association model uses gradient descent or maximum likelihood algorithm to obtain the model input parameters R p Number of outbound passengers input N I The best fit of .

5. The device for obtaining the number of outbound passengers in a vertical passenger elevator according to claim 1, characterized in that: The passenger number integration module M3 The predicted number of passengers N p Passenger input number N I Perform weighted calculation to get the number of passengers called out by the elevator N ; The weighted calculation step comprises: 1) From the passenger number input module M1 Get the number of elevator outbound passenger inputs N I ; 2) From the passenger number prediction module M2 Get the predicted number of elevator outbound passengers N p ; 3) According to the formula Calculate the number of passengers called out by the elevator N ;in, λ It is a weight factor determined according to the pressing duration input by the passenger.

6. An elevator group control strategy optimization system based on the vertical passenger elevator outbound call passenger number acquisition device according to any one of claims 1 to 5, characterized in that: It includes a car state calculation module, a call distribution module, a car control module and a call data and car state historical data storage module; The car status calculation module is responsible for integrating the number of passengers from the M3 Number of passengers receiving elevator calls N , and combined with the elevator control signal provided by the car control module to determine the current state of the elevator car, and transmit the relevant data to the call distribution module to ensure the coordination and efficiency of the elevator operation; The call distribution module is responsible for coordinating and controlling the entire elevator group control system, integrating the information of the car status calculation module and the number of passengers called by the elevator. N , generate elevator dispatch instructions and ensure that the instructions are executed correctly; The car control module performs specific car dispatching operations according to the dispatching strategy of the call distribution module. After execution, the module feeds back the operation results to the car status calculation module in real time to continuously optimize the response and dispatching strategy of the elevator. The call data and car status historical data storage module is responsible for data archiving and subsequent processing, recording the situation of each elevator dispatch and the output data of the passenger number integration module; the recorded data will be used to train and optimize the algorithm of the passenger prediction unit in the passenger number prediction module, continuously improving its prediction accuracy and the overall performance of the system.

Images