Elevator control method, elevator system and storage medium

Abstract

The invention discloses an elevator control method, an elevator system and a storage medium. The method comprises the steps that operation information of all elevators is obtained, and the operation information comprises stopping floors and the number of passengers; inputting the operation information into a pre-created prediction model for prediction to obtain a scheduling instruction; wherein the prediction model is obtained based on training of a time sequence model and a deep learning model; and at least one elevator is dispatched in response to the dispatching instruction. The technical problem that the elevator dispatching efficiency is low can be solved by collecting the operation data of each elevator in real time, predicting the passenger flow of the elevator by using the prediction model and generating the dispatching instruction, and then dispatching the elevator according to the dispatching instruction.

Description

Technical Field

[0001] The present application relates to the technical field of elevators, and in particular to an elevator control method, an elevator system and a storage medium. Background Art

[0002] With the continuous development of modern buildings and the increase in the number of high-rise buildings, elevators, as the main vertical transportation tool in buildings, their operating efficiency is directly related to the passengers' travel experience and the overall operational efficiency of the building.

[0003] However, when dealing with complex passenger flows and demands on different floors, the existing elevator control systems mostly use fixed logic or simple rules for scheduling, such as first-come-first-served, shortest waiting time, etc. However, these methods are difficult to achieve global optimization when faced with complex and dynamically changing passenger demands, resulting in low elevator scheduling efficiency and long waiting times for passengers.

[0004] Therefore, it is necessary to propose a method to improve the efficiency of elevator dispatching and solve the technical problem of low elevator dispatching efficiency. Summary of the invention

[0005] The main technical problem solved by the present application is to provide an elevator control method, an elevator system and a storage medium, which can solve the technical problem of low elevator dispatching efficiency.

[0006] In order to solve the above technical problems, a technical solution adopted by the present application is: the method is applied to an elevator system, the elevator system includes a plurality of elevators, and the method includes:

[0007] Obtaining the operation information of each elevator, wherein the operation information includes the stop floors and the number of passengers;

[0008] Input the operation information into a pre-created prediction model for prediction to obtain a scheduling instruction; wherein the prediction model is trained based on a time series model and a deep learning model;

[0009] In response to the dispatch instruction, at least one elevator is dispatched.

[0010] Optionally, the step of dispatching at least one elevator in response to the dispatch instruction comprises:

[0011] Parsing the dispatch instruction to determine a number of peak period information, wherein the peak period information includes a corresponding target floor and passenger demand;

[0012] According to the peak period information, the operation function of each elevator in the elevator system is determined to dispatch at least one elevator.

[0013] Optionally, after the step of dispatching at least one elevator in response to the dispatch instruction, the method further comprises:

[0014] According to the scheduling instruction, generating corresponding historical scheduling information;

[0015] The historical scheduling information is fed back to the prediction model for re-learning to obtain an updated prediction model.

[0016] Optionally, the process of creating the prediction model includes:

[0017] Get the historical elevator dataset;

[0018] Inputting the historical elevator data set into the time series model and the deep learning model respectively for prediction to obtain a first result and a second result;

[0019] According to the first result and the second result, a comprehensive index is obtained;

[0020] The pre-built initial prediction model is trained according to the comprehensive index to obtain the prediction model.

[0021] Optionally, the elevator system further includes a camera in the elevator, the elevator system communicates with an operator outside the elevator, and the step of obtaining the operation information of each elevator includes:

[0022] In response to the passenger reservation request sent by the operator outside the elevator, the reservation request is parsed to obtain the stop floor; and the image inside the elevator is collected by the camera inside the elevator, and face recognition is performed on the image inside the elevator to determine the number of passengers.

[0023] Optionally, the method further comprises:

[0024] Converting the image in the ladder into a digital signal;

[0025] Inputting the digital signal into a pre-created reference model for similarity matching to obtain a matching result;

[0026] If the matching result is that the image in the elevator contains abnormal human postures and / or restricted objects, an alarm mechanism is activated.

[0027] Optionally, the reference model stores reference features, and the step of inputting the digital signal into a pre-created reference model for similarity matching to obtain a matching result comprises:

[0028] Extracting digital features from the digital signal through the reference model;

[0029] Compare the digital feature with the reference feature and calculate the similarity;

[0030] When the similarity satisfies a preset similarity threshold, the matching result is that abnormal human postures and / or restricted objects exist in the image within the ladder.

[0031] Optionally, the operation information further includes passenger arrival time, and after the step of obtaining the operation information of each elevator, the step further includes:

[0032] Counting the arrival time of the passengers, determining the target stop time of the stop floor, and generating a stop notification;

[0033] The stop notification is returned to the elevator external operator.

[0034] In order to solve the above technical problems, another technical solution adopted by the present application is: an elevator system, the elevator system comprising:

[0035] A memory for storing executable program codes;

[0036] The processor is used to call and run the executable program code from the memory, so that the elevator system executes the elevator control method as described in any one of the above.

[0037] To solve the above technical problems, another technical solution adopted in the present application is: a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the elevator control method as described in any one of the above items is implemented.

[0038] Different from the prior art, an embodiment of the present application provides an elevator control method, which includes: obtaining operating information of each elevator, the operating information including the stop floors and the number of passengers; inputting the operating information into a pre-created prediction model for prediction to obtain a scheduling instruction; wherein the prediction model is trained based on a time series model and a deep learning model; and in response to the scheduling instruction, scheduling at least one elevator.

[0039] The beneficial effect of the embodiments of the present application is: by collecting the operating data of each elevator in real time, using a prediction model to predict the passenger flow of the elevator and generate dispatch instructions, and then dispatching the elevator according to the dispatch instructions, the technical problem of low elevator dispatching efficiency can be solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a flowchart of the first embodiment of the elevator control method of the present application;

[0041] Figure 2 It is a flow chart of the second embodiment of the elevator control method of the present application;

[0042] Figure 3It is a flowchart of the third embodiment of the elevator control method of the present application;

[0043] Figure 4 2 is a schematic diagram of the elevator system composition of the third embodiment of the elevator control method of the present application;

[0044] Figure 5 is an exemplary structural block diagram of an elevator system of the elevator control method of the present application;

[0045] Figure 6 It is an exemplary structural block diagram of a computer-readable storage medium of the elevator control method of the present application. DETAILED DESCRIPTION

[0046] The present application is described in detail below with reference to the accompanying drawings and implementation methods.

[0047] See also Figure 1 , Figure 1 It is a flow chart of the first embodiment of the elevator control method of the present application. It should be noted that if there is substantially the same result, the method of the present application is not limited to Figure 1 The process sequence shown is limited. Figure 1 As shown, the method is applied to an elevator system, the elevator system includes a plurality of elevators, and the method includes:

[0048] Step S110, obtaining the operation information of each elevator, wherein the operation information includes the stop floors and the number of passengers;

[0049] This embodiment takes into account that traditional elevator systems generally use fixed logic or simple rules for scheduling, such as first-come-first-served, shortest waiting time, etc. These methods are difficult to achieve global optimization when facing complex and dynamically changing passenger demands, resulting in low elevator scheduling efficiency and long passenger waiting time; at the same time, the lack of real-time analysis capabilities of elevator operation data and passenger demand makes it difficult for the elevator system to quickly respond to passengers' real-time needs, further exacerbating congestion and waiting problems during peak hours; and due to the limitations of the scheduling algorithm, passengers often need to wait for a long time before they can take the elevator, especially during rush hours and when demand for specific floors is high, and the passenger experience is poor; therefore, this embodiment can analyze the daily operation data of the elevator to infer the peak period and make scheduling decisions in a timely manner.

[0050] Specifically, sensors and artificial intelligence (AI) cameras can be used to monitor the operating status of each elevator in real time. The collected operating information can include the stop floors and the number of passengers of each elevator. The stop floors can be the floors where each elevator needs to stop; the number of passengers can be the number of passengers corresponding to each stop floor in the elevator.

[0051] Step S120, inputting the operation information into a pre-created prediction model for prediction to obtain a scheduling instruction; wherein the prediction model is trained based on a time series model and a deep learning model;

[0052] Specifically, the collected real-time operation information is input into the prediction model for prediction, and the elevator dispatch instructions are obtained. Among them, this prediction model combines the time series model and the deep learning model. By combining historical data and real-time data, passenger demand can be more accurately predicted and elevator dispatch can be optimized. The operating status of the elevator can be analyzed, passenger demand can be predicted, and optimized dispatch instructions can be generated. The time series model can be used to analyze and predict data that changes over time; the deep learning model can be used to process and analyze complex data using neural networks.

[0053] More specifically, the prediction model can first use the time series model to analyze the time trend of elevator operation data, such as the changes in passenger flow during peak and off-peak hours; the deep learning model further processes the changes in passenger flow to identify different characteristics and patterns corresponding to the changes in passenger flow. Then, the model integrates this information to generate a dispatch instruction for at least one elevator in the elevator system.

[0054] Step S130, dispatching at least one elevator in response to the dispatch instruction.

[0055] Specifically, the dispatching instructions are generated according to the current passenger flow changes, so that the elevators in the elevator system can adapt to the passenger flow changes to stop at the corresponding floors, so that they can operate according to the optimization plan. The dispatching instructions may include adjusting the running path and the stop floors of the elevators to minimize the waiting time of passengers and the empty operation of the elevators.

[0056] For example, an elevator in the elevator system may be instructed to stop at a specific floor for a longer time to wait for an expected large number of passengers to enter, or to give priority to high-traffic floors during certain periods. In addition, the system can also make dynamic adjustments based on real-time data, such as quickly dispatching the nearest elevator to a floor when a sudden large number of passengers appears on that floor.

[0057] This embodiment collects the operating data of each elevator in real time, uses a prediction model to predict the passenger flow of the elevator and generate a dispatch instruction, and then dispatches the elevator according to the dispatch instruction, which can effectively improve the efficiency of elevator dispatch.

[0058] See also Figure 2 , Figure 2 1 is a flow chart of the second embodiment of the elevator control method of the present application. In this embodiment, the elevator system further includes a camera inside the elevator, the elevator system communicates with an operator outside the elevator, the reference model stores reference features, and the method includes the following steps:

[0059] Step S201, in response to the passenger reservation request sent by the elevator external operator, parsing the reservation request to obtain the stop floor;

[0060] This embodiment takes into account that in the existing elevator system, in the scenario of multiple elevators in an office building, during peak hours, passengers may need to wait for a long time before using the elevator, which wastes passengers' time and property electricity. At the same time, traditional elevators only interact with the property through floor buttons or intercom systems. There may be situations where people are trapped but the signal in the elevator is poor and it is difficult to communicate with the outside world, and the property is unattended or not noticed, resulting in people being trapped for a long time and the emergency situation is difficult to handle.

[0061] Therefore, this embodiment proposes a method for realizing intelligent dispatching and efficient and safe operation of elevators based on AI cameras, and realizes functions such as intelligent dispatching, enhanced safety, energy saving and environmental protection by integrating advanced sensors, control algorithms and communication technologies.

[0062] Specifically, the elevator external operator may be a control device located outside the elevator, and may include at least one of a button, a touch screen, and a voice recognition system, and is used for passengers to make a call or reserve a destination floor and a time to arrive at the elevator door before entering the elevator, and the passenger reservation request may be a request made by the passenger through the elevator external operator. When the passenger sends a reservation request through the elevator external operator outside the elevator, the system receives the request and parses the information contained therein, and the information may be, for example, the stop floor that the passenger wishes to go to.

[0063] Step S202, collecting images in the elevator through the camera in the elevator, and performing face recognition on the images in the elevator to determine the number of passengers.

[0064] Specifically, the camera in the elevator can be an AI camera installed in the elevator, used to capture images inside the elevator; the images inside the elevator can be photos of passengers taken by the camera in the elevator; face recognition can be a method of determining the identity and number of passengers by analyzing and identifying facial features in the images inside the elevator.

[0065] When passengers enter the elevator, the camera inside the elevator will capture real-time images of the elevator, and the system uses facial recognition technology to analyze these images to determine the number of passengers currently in the elevator. Facial recognition technology can distinguish the facial features of each passenger and calculate the total number of people in the elevator on each floor.

[0066] Step S203, inputting the operation information into a pre-created prediction model for prediction to obtain a scheduling instruction; wherein the prediction model is trained based on a time series model and a deep learning model;

[0067] The above step S120 may be performed and will not be described in detail here.

[0068] Step S204, parsing the dispatch instruction to determine some peak period information, wherein the peak period information includes corresponding target floors and passenger demands;

[0069] Specifically, by parsing the dispatch instructions generated by the prediction model, relevant information about the peak period can be extracted from the dispatch instructions. The peak period information can include the target floors with the largest passenger demand during these periods, as well as the passenger demand conditions on these floors.

[0070] For example, the dispatch instruction can describe the passenger flow trend during peak hours, for example, the passenger demand for certain office floors surges between 8 and 9 a.m. The target floors can include major office floors, shopping mall floors, or dining areas, and the passenger demand can be the intensity and number of passengers who want to quickly reach these floors. By analyzing this information, the system can more accurately understand the characteristics of peak hours and the distribution of passenger demand.

[0071] Step S205: determining the operation function of each elevator in the elevator system according to the peak period information to dispatch at least one elevator.

[0072] Specifically, based on the peak period information obtained through analysis, the specific operating function of each elevator during the peak period can be determined to ensure that passenger needs can be efficiently met during these periods.

[0073] The system will adjust the operation strategy of each elevator based on the peak information. For example, during peak hours, more elevators can be arranged to give priority to target floors with high demand, reducing the waiting time for passengers on these floors. For example, the operation function can include a directional service function, which can specify that certain elevators only serve specific high-demand floors during peak hours, such as only traveling between the first floor and the office floor; it can also include a priority stop function, which adjusts the stop strategy of the elevator to give priority to the target floors with high demand during peak hours.

[0074] As a result, the elevator system can operate efficiently during peak hours, minimize passengers' waiting time, and improve the overall passenger experience; at the same time, the operation strategy can also be dynamically adjusted based on real-time data to ensure a smooth transition between peak and off-peak hours and maintain the efficient operation of the elevator system.

[0075] Step S206, generating corresponding historical scheduling information according to the scheduling instruction;

[0076] Specifically, this embodiment can record the process and results of executing the dispatch instruction and generate historical dispatch information for subsequent analysis and optimization of the prediction model. The historical dispatch information can include detailed data such as the execution status of each dispatch, passenger waiting time, number of passengers, elevator stop floors, etc.

[0077] Step S207, feeding back the historical scheduling information to the prediction model for re-learning to obtain an updated prediction model.

[0078] Specifically, the generated historical scheduling information is fed back to the prediction model, so that the prediction model analyzes and trains the fed-back historical scheduling information through the re-learning process, and updates its own prediction ability and accuracy. Among them, re-learning is a continuous process. By continuously inputting new historical scheduling data, the prediction model can gradually optimize itself, improve its ability to handle complex situations and the accuracy of scheduling decisions.

[0079] Furthermore, the creation process of the prediction model includes:

[0080] Step S210, obtaining a historical elevator data set;

[0081] Specifically, the historical operation data of the elevator system is collected and sorted to form a historical elevator data set. The historical elevator data set may contain the operation information of all elevators in the past period of time, including but not limited to: the record of the elevator's stop floors; the number of passengers at each stop; the elevator's operating time and waiting time; the operation data of different time periods (such as morning peak, evening peak, working hours, rest time); elevator failure and maintenance records.

[0082] Step S220, inputting the historical elevator data set into the time series model for processing to obtain a first result; inputting the historical elevator data set into the deep learning model for processing to obtain a second result;

[0083] Specifically, time series models can be used to analyze the time trends and periodic changes of elevator operation data for prediction; deep learning models can process complex historical data and identify deep patterns and features in the data for prediction.

[0084] Step S230, obtaining a comprehensive index according to the first result and the second result;

[0085] Specifically, the first result output by the time series model and the second result output by the deep learning model are combined, and a comprehensive index is calculated through a certain evaluation method to measure the accuracy and effectiveness of the predictions of the two models.

[0086] Among them, the comprehensive indicators may include at least one of the following: prediction error, which is used to evaluate the error between the predicted values ​​of the two models and the actual values, such as mean square error, mean absolute error, etc.; accuracy, which is used to measure the accuracy of model predictions, such as accuracy rate and precision; flexibility, which is used to evaluate the adaptability and stability of the model under different situations; comprehensive score, which is used to combine multiple evaluation indicators into a total score through weighted average or other methods to evaluate the overall performance of the model.

[0087] Step S240, training the pre-built initial prediction model according to the comprehensive index to obtain the prediction model.

[0088] Specifically, according to the evaluation results of the comprehensive indicators, the initial prediction model can be optimized iteratively so that the initial prediction model can combine the advantages of the time series model and the deep learning model to generate more accurate and reliable prediction results, and finally obtain the prediction model. Among them, the initial prediction model can be constructed by fusing the time series model and the deep learning model through weighted averaging, stacking model, ensemble learning and other methods.

[0089] In addition, after training is completed, the prediction model can be verified using an independent validation dataset to ensure the stability and reliability of its predictive ability.

[0090] In this embodiment, the passenger operates the external operator of the elevator to initiate a request for reserving a stopping floor, and the AI ​​camera in the elevator obtains the number of passengers, and then inputs the stopping floors and passengers into the prediction model for prediction, and generates corresponding dispatch instructions. The peak period information can be determined according to the dispatch instructions, and at least one elevator can be dispatched according to the peak period information. This can effectively realize adaptive elevator scheduling while improving the convenience of passenger operation.

[0091] See also Figure 3 , Figure 3 This is a flow chart of the third embodiment of the elevator control method of the present application. Figure 2 and Figure 3 In the embodiment shown, in this embodiment, the method includes the following steps:

[0092] Step S310, converting the image in the elevator into a digital signal;

[0093] This embodiment takes into account the safety precautions of elevators, such as the situation of carrying chargers or electric vehicles in the elevator, and dangerous situations such as passengers falling and children riding the elevator alone. Traditional elevators are difficult to identify the above scenarios, which lead to tragedies. Therefore, this embodiment proposes a method based on AI cameras to achieve intelligent dispatching and efficient and safe operation of elevators, and realizes functions such as safety enhancement by integrating advanced sensors, control algorithms and communication technologies.

[0094] Specifically, the images inside the elevator captured by the camera inside the elevator are transmitted in the form of analog signals. In order to calculate and analyze the image data in subsequent steps, the images inside the elevator can be converted into digital signals, including pixel values, color information, and timestamps.

[0095] Step S320, extracting digital features from the digital signal through the reference model;

[0096] Specifically, the reference model can be a convolutional neural network, etc., which is used to perform convolution operations on the input digital signal to extract digital features that represent the image, such as edge digital features, texture digital features, shape digital features, etc.

[0097] Step S330, comparing the digital feature with the reference feature to calculate the similarity;

[0098] Specifically, similarity calculation can adopt methods such as cosine similarity and Euclidean distance. Reference features can be predefined based on a large amount of image data, and can include feature sets representing normal and abnormal states. By comparing digital features with reference features, it can be determined whether the current image has abnormal conditions.

[0099] Step S340: When the similarity meets a preset similarity threshold, the matching result is that abnormal human postures and / or restricted objects exist in the image in the ladder.

[0100] Specifically, the similarity threshold can be used to distinguish between normal and abnormal situations in the elevator. If the calculated similarity exceeds the preset similarity threshold, it means that the current image feature is highly similar to the abnormal reference feature, and it is determined that there are abnormal human postures and / or restricted objects in the elevator. Among them, the abnormal human posture can be abnormal postures such as fainting and falling; the restricted objects can be electric vehicles, pets, flammable and explosive items, etc.

[0101] Step S350: If the matching result is that abnormal human postures and / or restricted objects exist in the image in the elevator, an alarm mechanism is activated.

[0102] Specifically, when the matching result indicates that there is an abnormal or dangerous situation, the alarm mechanism of the elevator system is triggered. The alarm mechanism can include various forms such as sound alarm, light alarm, and notification of security personnel. The system can sound an alarm through the speaker in the elevator and send an emergency notification through the control center outside the elevator to ensure that timely action can be taken to deal with abnormal or dangerous situations.

[0103] Furthermore, the operation information also includes the passenger arrival time, and after step S110, it also includes:

[0104] The passenger arrival time is counted, the target stop time of the stop floor is determined, and a stop notification is generated; and the stop notification is returned to the elevator external operator.

[0105] Specifically, the target stop time can be the expected stop time of the elevator at the floor, ensuring that passengers have enough time to get in and out of the elevator. The stop notification can be used to inform passengers that the elevator is about to arrive at the target floor and the expected target stop time, so that passengers can prepare themselves.

[0106] This embodiment converts the image in the elevator captured by the camera in the elevator into a digital signal, then extracts the features of the digital signal through the reference model, and calculates the similarity based on the preset reference features. When the similarity meets the preset threshold, it determines that the matching result is an abnormal situation in the elevator, and then activates the alarm, which can improve the safety of elevator operation. At the same time, the generated stop notification is displayed or broadcasted through the operator outside the elevator to ensure that passengers can get the information in time before the elevator arrives, which can not only improve the passenger experience, but also further optimize the operation efficiency of the elevator.

[0107] See also Figure 4 , Figure 4 Schematic diagram of the elevator system composition of the third embodiment of the elevator control method of the present application. Figure 4 As shown, the elevator system may include an AI camera inside the elevator (i.e., a camera inside the elevator), a camera outside the elevator, an operator inside the elevator, an operator outside the elevator, and a central control system.

[0108] Among them, the AI ​​camera in the elevator is used to monitor and record the situation in the elevator, realize passenger counting, alarm for people fainting or falling, and alarm for electric vehicles or pets entering the elevator; the operator in the elevator can integrate communication, audio, button and display modules for floor selection and communication with external operators or property master controllers; the camera outside the elevator is used for traditional monitoring and recording the situation in the elevator, and counting the number of passengers; the operator outside the elevator can integrate communication, audio, button and display modules for floor selection and reservation operations, and communicate with external operators or property master controllers.

[0109] Therefore, the central control system of the elevator system can dynamically adjust the elevator operation plan according to the total number of elevators, the number of passengers in the elevator, the demand and the elevator operation status during peak hours, which can save time; during non-peak hours, the number of operations can be reasonably reduced and the lights inside the elevator can be turned off to save resources. Among them, the most reasonable scheduling is calculated by counting the number of passengers and demand on each floor and combining the current elevator operation status; by analyzing the daily elevator operation data, the peak period is inferred and the scheduling decision is made; and the sleep time can be set and some functions can be hibernated to save energy.

[0110] At the same time, the AI ​​camera in the elevator can identify electric vehicles, electric vehicle batteries and human posture to alarm or suspend the operation of the elevator; by converting images into digital signals, processing and analyzing the digital signals to calculate the corresponding similarity; when the similarity meets the preset similarity threshold, the elevator system can alarm passengers and / or property to suspend the operation of the elevator.

[0111] In addition, passengers can also communicate with relevant floors through voice through the elevator controller and the elevator external controller; by using the elevator external controller, stair reservation operations can be realized through QR code or mini program, saving passengers waiting time.

[0112] See also Figure 5 , Figure 5 is an exemplary structural block diagram of an elevator system of the elevator control method of the present application. Figure 5 As shown, the elevator system 500 of the present application may include a processor 501 and a memory 502, wherein the processor 501 and the memory 502 communicate with each other via a bus. The memory 502 stores program instructions for elevator control, and when the program instructions are executed by the processor 501, the method steps described in the above text in conjunction with the accompanying drawings are implemented:

[0113] Obtaining the operation information of each elevator, wherein the operation information includes the stop floors and the number of passengers;

[0114] Input the operation information into a pre-created prediction model for prediction to obtain a scheduling instruction; wherein the prediction model is trained based on a time series model and a deep learning model;

[0115] In response to the dispatch instruction, at least one elevator is dispatched.

[0116] Optionally, the processor 501 executes the step of dispatching at least one elevator in response to the dispatch instruction, including:

[0117] Parsing the dispatch instruction to determine a number of peak period information, wherein the peak period information includes a corresponding target floor and passenger demand;

[0118] According to the peak period information, the operation function of each elevator in the elevator system is determined to dispatch at least one elevator.

[0119] Optionally, after the processor 501 executes the step of dispatching at least one elevator in response to the dispatch instruction, the step further includes:

[0120] According to the scheduling instruction, generating corresponding historical scheduling information;

[0121] The historical scheduling information is fed back to the prediction model for re-learning to obtain an updated prediction model.

[0122] Optionally, the processor 501 executes the process of creating the prediction model including:

[0123] Get the historical elevator dataset;

[0124] Inputting the historical elevator data set into the time series model and the deep learning model respectively for prediction to obtain a first result and a second result;

[0125] According to the first result and the second result, a comprehensive index is obtained;

[0126] The pre-built initial prediction model is trained according to the comprehensive index to obtain the prediction model.

[0127] Optionally, the processor 501 executes the elevator system further including a camera in the elevator, the elevator system communicates with an operator outside the elevator, and the step of obtaining the operation information of each elevator includes:

[0128] In response to the passenger reservation request sent by the operator outside the elevator, the reservation request is parsed to obtain the stop floor; and the image inside the elevator is collected by the camera inside the elevator, and face recognition is performed on the image inside the elevator to determine the number of passengers.

[0129] Optionally, the processor 501 executing the method further includes:

[0130] Converting the image in the ladder into a digital signal;

[0131] Inputting the digital signal into a pre-created reference model for similarity matching to obtain a matching result;

[0132] If the matching result is that the image in the elevator contains abnormal human postures and / or restricted objects, an alarm mechanism is activated.

[0133] Optionally, the processor 501 executes the step of storing reference features in the reference model, inputting the digital signal into a pre-created reference model for similarity matching, and obtaining a matching result, including:

[0134] Extracting digital features from the digital signal through the reference model;

[0135] Compare the digital feature with the reference feature and calculate the similarity;

[0136] When the similarity satisfies a preset similarity threshold, the matching result is that abnormal human postures and / or restricted objects exist in the image within the ladder.

[0137] Optionally, after the processor 501 executes the step of obtaining the operation information of each elevator, the operation information further includes the passenger arrival time, and further includes:

[0138] Counting the arrival time of the passengers, determining the target stop time of the stop floor, and generating a stop notification;

[0139] The stop notification is returned to the elevator external operator.

[0140] See also Figure 6 , Figure 6 is an exemplary structural block diagram of a computer-readable storage medium of the elevator control method of the present application. Figure 6 As shown, a computer program 601 is stored in the computer-readable storage medium 600. When the computer program 601 is executed on a computer, the computer executes the above-mentioned related method steps to implement an elevator control method in the above-mentioned embodiment.

[0141] The above scheme obtains the operation information of each elevator, including the stop floors and the number of passengers; inputs the operation information into a pre-created prediction model for prediction, and obtains a dispatch instruction; wherein the prediction model is trained based on a time series model and a deep learning model; and dispatches at least one elevator in response to the dispatch instruction. By collecting the operation data of each elevator in real time, using the prediction model to predict the passenger flow of the elevator and generate a dispatch instruction, and then dispatching the elevator according to the dispatch instruction, the technical problem of low elevator dispatch efficiency can be solved.

[0142] In the several embodiments provided in the present application, it should be understood that the disclosed methods, elevator systems and storage media can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0143] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0144] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0145] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be an elevator system, a personal computer, a server, or a network device, etc.) or a processor (processor) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0146] The above description is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process change made by using the contents of the present application specification and drawings is

[0147] The invention may be directly or indirectly used in other related technical fields, and shall be included in the patent protection scope of this application.

Claims

1. An elevator control method, characterized in that: The method is applied to an elevator system, the elevator system comprising a plurality of elevators, and the method comprises: Obtaining the operation information of each elevator, wherein the operation information includes the stop floors and the number of passengers; Input the operation information into a pre-created prediction model for prediction to obtain a scheduling instruction; wherein the prediction model is trained based on a time series model and a deep learning model; In response to the dispatch instruction, at least one elevator is dispatched.

2. The method according to claim 1, characterized in that The step of dispatching at least one elevator in response to the dispatch instruction comprises: Parsing the dispatch instruction to determine a number of peak period information, wherein the peak period information includes a corresponding target floor and passenger demand; According to the peak period information, the operation function of each elevator in the elevator system is determined to dispatch at least one elevator.

3. The method according to claim 1, characterized in that After the step of dispatching at least one elevator in response to the dispatch instruction, the method further includes: Generate corresponding historical scheduling information according to the scheduling instruction; The historical scheduling information is fed back to the prediction model for re-learning to obtain an updated prediction model.

4. The method according to claim 1, characterized in that: The process of creating the prediction model includes: Get the historical elevator dataset; Inputting the historical elevator data set into the time series model and the deep learning model respectively for prediction to obtain a first result and a second result; According to the first result and the second result, a comprehensive index is obtained; The pre-built initial prediction model is trained according to the comprehensive index to obtain the prediction model.

5. The method according to claim 1, characterized in that: The elevator system further includes a camera in the elevator, the elevator system communicates with an operator outside the elevator, and the step of obtaining the operation information of each elevator includes: In response to the passenger reservation request sent by the elevator external operator, parsing the reservation request to obtain the stop floor; and The elevator camera collects images inside the elevator, and performs face recognition on the images inside the elevator to determine the number of passengers.

6. The method according to claim 5, characterized in that The method further comprises: Converting the image in the ladder into a digital signal; Inputting the digital signal into a pre-created reference model for similarity matching to obtain a matching result; If the matching result is that the image in the elevator contains abnormal human postures and / or restricted objects, an alarm mechanism is activated.

7. The method according to claim 6, characterized in that The reference model stores reference features, and the step of inputting the digital signal into the pre-created reference model for similarity matching to obtain a matching result comprises: Extracting digital features from the digital signal through the reference model; Compare the digital feature with the reference feature and calculate the similarity; When the similarity satisfies a preset similarity threshold, the matching result is that abnormal human postures and / or restricted objects exist in the image within the ladder.

8. The method according to claim 5, characterized in that The operation information also includes the passenger arrival time. After the step of obtaining the operation information of each elevator, the step further includes: Counting the arrival time of the passengers, determining the target stop time of the stop floor, and generating a stop notification; The stop notification is returned to the elevator external operator.

9. An elevator system, characterized in that: The elevator system comprises: A memory for storing executable program codes; A processor is used to call and run the executable program code from the memory, so that the elevator system executes the elevator control method according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the elevator control method according to any one of claims 1 to 8 is implemented.

Images