Novel elevator multi-scene operation intelligent control method and system based on AI model

Abstract

The invention discloses a novel elevator multi-scene operation intelligent control method and system based on an AI model. The novel elevator multi-scene operation intelligent control method comprises the steps that the elevator operation speed, the car load capacity, the environment temperature and humidity, the steel rope dynamic deformation quantity and floor call distribution data are collected in real time; on the basis of an AI scene recognition model, the type of an operation scene where the elevator is located is recognized according to the parameters so as to match a corresponding AI regulation and control strategy, the brake pressure of the elevator is calculated through a brake pressure regulation and control model, and the response time of the elevator is calculated according to a response time regulation and control model; and the maximum springback distance when the elevator stops stably is obtained, a compensation factor is calculated in combination with the operation parameters of the elevator, the calculated brake pressure and response time are updated, and the updated parameters are used for conducting self-adaptive adjustment on the operation parameters of the elevator. Compared with the prior art, comprehensive data are collected, the scene where the elevator is located can be accurately recognized to be matched with the AI regulation and control strategy, accurate control is provided for safe operation of the elevator, and the elevator taking experience of a user is improved while energy consumption is reduced.

Description

Technical Field

[0001] The present invention relates to the field of elevator control technology, and in particular to a novel intelligent control method and system for multi-scenario operation of elevators based on an AI model. Background Art

[0002] In today's elevator operation control field, traditional technologies have exposed many significant shortcomings. First, most traditional elevator operation control methods struggle to accurately and comprehensively collect elevator operating parameters in real time. They often focus only on basic data such as elevator speed and load, while ignoring the impact of other data on elevator operation. This results in limited control of the elevator's overall operating status and limited accuracy, making it impossible to rationally adjust elevator operation strategies based on comprehensive data.

[0003] On the other hand, traditional elevators have serious deficiencies in matching scene recognition with control strategies. They are unable to accurately identify the type of operating scenario in real time based on the elevator's operating status, relying solely on fixed, rigid operating modes. For example, during peak congestion, elevator operation cannot be effectively optimized, resulting in long passenger wait times and increased energy consumption. During off-peak energy-saving scenarios, operating parameters cannot be adjusted promptly to reduce energy consumption. In emergency braking scenarios, traditional control methods lack the precision to adjust response speed and brake pressure, failing to ensure passenger safety. Even in normal, stable conditions, further optimization of operating parameters to enhance passenger comfort and equipment life is difficult. Summary of the Invention

[0004] In order to solve at least one of the technical problems raised above, the present invention provides a new intelligent control method and system for multi-scenario operation of elevators based on an AI model.

[0005] In a first aspect, the present invention provides a novel intelligent control method for multi-scenario operation of an elevator based on an AI model, the method comprising:

[0006] Real-time collection of elevator operating parameters, including elevator speed, car load, ambient temperature and humidity, dynamic deformation of steel ropes, and floor call distribution data;

[0007] Based on the pre-trained AI scene recognition model, the elevator's operating scenario type is identified according to the current operating parameters. The scenario types include peak congestion scenarios, off-peak energy-saving scenarios, emergency braking scenarios, and normal stable scenarios.

[0008] The AI control strategy is matched to the identified scenario type, the brake pressure of the elevator is calculated using the brake pressure control model, and the response time of the elevator is calculated using the response time control model.

[0009] The maximum rebound distance of the elevator when it stops is obtained, and a compensation factor is calculated based on the maximum rebound distance and the elevator's operating parameters. The calculated brake pressure and response time are updated using the brake compensation factor, and the updated brake pressure and response time are used to adaptively adjust the elevator's operating parameters.

[0010] Preferably, the brake pressure control model building process includes:

[0011] Calculate a first relationship function of the influence of speed and acceleration on brake pressure based on the elevator acceleration, the influence weight of speed on brake pressure and the influence weight of brake pressure;

[0012] Calculate a second relationship function of the effects of temperature and humidity on brake pressure based on the reference temperature and humidity and the actual temperature and humidity collected in real time during elevator operation;

[0013] The static load and car mass of the elevator are obtained, and a brake pressure control model is constructed according to the static load, car mass, the first relationship function and the second relationship function.

[0014] Preferably, the process of constructing the response time control model includes:

[0015] Calculate the temperature difference between the reference temperature and the actual temperature, and the humidity difference between the reference humidity and the actual humidity respectively;

[0016] The maximum allowable response time of the elevator operation is obtained, and a response time control model is constructed based on the maximum allowable response time, acceleration, speed, temperature difference, and humidity difference of the elevator operation.

[0017] Preferably, the calculating of the compensation factor according to the maximum rebound distance and the operating parameters of the elevator, and updating the calculated brake pressure and response time using the brake compensation factor, comprises:

[0018] Calculate the compensation factor based on the elevator's maximum rebound distance, speed, and speed sensitivity coefficient;

[0019] Multiplying the calculated brake pressure by the compensation factor to obtain an updated brake pressure;

[0020] Divide the calculated response time by the compensation factor to obtain the updated response time.

[0021] Preferably, the pre-trained AI scene recognition model identifies the type of operation scene of the elevator according to the current operation parameters, including:

[0022] Determine whether the actual temperature collected in real time during elevator operation exceeds a first preset value. When the temperature exceeds the first preset value, obtain the load in the elevator car and identify the type of operation scenario of the elevator based on the load.

[0023] Preferably, after obtaining the load in the elevator car, the method further includes matching the target stop floor according to the load, including:

[0024] When the load is greater than a second preset value, the floor closest to the current position is used as the target docking floor;

[0025] When the load is less than or equal to a second preset value, the median floor of the called floors in the elevator car is used as the target stopping floor.

[0026] Preferably, the method further comprises adopting different execution logics for different operation scenario types, including:

[0027] In peak congestion scenarios, priority is given to shortening response time and increasing braking force, while simultaneously activating the car pre-dispatching algorithm to reduce inter-floor dwellings;

[0028] In low-peak energy-saving scenarios, a segmented soft braking strategy is adopted, combined with coasting to reduce motor energy consumption;

[0029] In emergency braking scenarios, redundant braking units are triggered, and the braking trajectory is simulated through the digital twin system to avoid resonance risks.

[0030] In a second aspect, the present invention provides a novel elevator multi-scenario operation intelligent control system based on an AI model, the system comprising:

[0031] Data acquisition unit, used to collect real-time operating parameters of the elevator, including elevator speed, car load, ambient temperature and humidity, dynamic deformation of the steel rope, and floor call distribution data;

[0032] A scene recognition unit is used to identify the type of operating scene the elevator is in based on the pre-trained AI scene recognition model and the current operating parameters. The scene types include peak congestion scenes, off-peak energy-saving scenes, emergency braking scenes, and normal stable scenes;

[0033] The strategy matching unit is used to match the corresponding AI control strategy according to the identified scenario type, calculate the elevator's brake pressure through the brake pressure control model, and calculate the elevator's response time based on the response time control model;

[0034] The parameter adjustment unit is used to obtain the maximum rebound distance when the elevator stops, calculate the compensation factor based on the maximum rebound distance and the elevator's operating parameters, use the brake compensation factor to update the calculated brake pressure and response time, and use the updated brake pressure and response time to adaptively adjust the elevator's operating parameters.

[0035] In a third aspect, the present invention also provides an electronic device comprising a processor and a memory, wherein the memory is used to store computer program code, and the computer program code comprises computer instructions. When the processor executes the computer instructions, the electronic device executes the method as described in the first aspect above and any possible implementation thereof.

[0036] In a fourth aspect, the present invention also provides a computer-readable storage medium, in which a computer program is stored. The computer program includes program instructions, and when the program instructions are executed by a processor of an electronic device, the processor executes the method as described in the first aspect above and any possible implementation method thereof.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] The novel AI-based intelligent control method for multi-scenario elevator operation, provided by this invention, effectively addresses these technical challenges. This method comprehensively and in real time collects data on elevator speed, car load, ambient temperature and humidity, dynamic rope deformation, and floor call distribution, providing a rich and accurate data foundation for subsequent intelligent decision-making. Leveraging a pre-trained AI scenario recognition model, it can accurately identify various elevator operating scenarios, including peak congestion scenarios, off-peak energy-saving scenarios, emergency braking scenarios, and normal stable scenarios. Corresponding AI control strategies are matched to different scenario types, and with the help of brake pressure control models and response time control models, the required brake pressure and response time are accurately calculated to achieve precise control. Furthermore, by obtaining the maximum rebound distance when the elevator comes to a complete stop and calculating a compensation factor based on the elevator's operating parameters, the brake pressure and response time are updated, thereby achieving adaptive adjustment of the elevator's operating parameters.

[0039] At the same time, in actual applications, if in peak congestion scenarios, after accurately identifying the scenario through analysis of floor call distribution data, the corresponding AI control strategy is used to rationally plan the elevator's running path and stop floors, which can significantly reduce passenger waiting time and improve elevator efficiency. In low-peak energy-saving scenarios, according to parameters such as ambient temperature and humidity and the results of low-peak scenario identification, the elevator's running speed, lighting and other parameters are automatically adjusted to reduce elevator energy consumption. In the face of emergency braking scenarios, the AI model is used to quickly and accurately identify and calculate the appropriate braking pressure through the brake pressure control model. Combined with the response time control model, the elevator can brake smoothly in the shortest time, greatly improving passenger safety. In normal and stable scenarios, the operating status is continuously fine-tuned according to the elevator operating parameters and compensation factors to reduce the wear of elevator components and extend the service life of the equipment. According to statistics, the wear rate of related components has been reduced by more than 15%, effectively reducing maintenance costs.

[0040] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the background technology, the drawings required for use in the embodiments of the present invention or the background technology will be described below.

[0042] The accompanying drawings herein are incorporated into and constitute a part of the specification. These drawings illustrate embodiments consistent with the present disclosure and, together with the specification, are used to explain the technical solutions of the present disclosure.

[0043] Figure 1 A flowchart of a novel intelligent control method for multi-scenario operation of an elevator based on an AI model provided by an embodiment of the present invention;

[0044] Figure 2 A schematic diagram of the structure of a new elevator multi-scenario operation intelligent control system based on an AI model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0046] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present invention. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0047] See also Figure 1 , Figure 1 The following is a flow chart of a novel intelligent control method for multi-scenario operation of elevators based on an AI model provided by an embodiment of the present invention. Figure 1 As shown, the method includes:

[0048] S10, real-time collection of elevator operating parameters, including elevator operating speed, car load, ambient temperature and humidity, dynamic deformation of steel ropes, and floor call distribution data;

[0049] Elevator speed refers to the speed of the elevator car during operation, typically measured in meters per second. It reflects the speed of the elevator and is a key indicator of its efficiency and comfort. This speed is typically acquired by installing a speed sensor on the elevator's drive system or car. A common speed sensor is a rotary encoder, which connects to the moving parts of the elevator's traction motor or car. As the parts rotate, they generate pulse signals, and the number of pulses per unit time is used to calculate the elevator's speed.

[0050] Car load capacity refers to the weight carried within the elevator car, measured in kilograms or tons. This parameter is used to monitor elevator overload and ensure safe operation. Load cells are typically installed at the base of the elevator car or in the suspension system. When the car is loaded, the load cell experiences compression or tension, causing the strain gauge within it to deform, resulting in a change in resistance. This change in resistance is then used to calculate the actual weight carried by the car.

[0051] Ambient temperature refers to the air temperature inside an elevator car or related environment, such as the machine room, and is measured in degrees Celsius (°C). Ambient humidity refers to the water vapor content in the air, usually expressed as relative humidity (%). Temperature and humidity have a certain impact on the elevator's electrical equipment and mechanical components, as well as passenger comfort. Temperature and humidity sensors can be used to collect data. Temperature and humidity sensors are generally digital or analog and can be installed in elevator cars, machine rooms, hoistways, and other locations where temperature and humidity need to be monitored. The sensors sense changes in temperature and humidity in the environment, converting physical quantities into electrical signals. Signal processing circuits then transmit the data to the acquisition system.

[0052] During elevator operation, the steel rope undergoes minute deformations due to the weight of the car, the pull of the traction machine, and various operating forces. The dynamic deformation of the steel rope refers to these time-varying deformations, which reflect the stress condition and fatigue level of the steel rope. This can be measured using either fiber Bragg grating (FBG) sensors or strain gauge sensors. Fiber Bragg grating (FBG) sensors attach a fiber Bragg grating (FBG) to the surface of the steel rope. When the rope deforms, the reflected wavelength of the FBG changes, and the rope deformation is detected by detecting this wavelength change. Strain gauge sensors attach a strain gauge to the steel rope. When the rope deforms under stress, the resistance of the strain gauge changes, and the rope deformation is calculated by measuring this resistance change.

[0053] Floor call distribution data includes information such as the number of times the elevator call button is pressed for each floor, the duration of the press, and the direction of the call. This data is used to analyze elevator usage frequency and passenger flow distribution, thereby optimizing elevator operation and scheduling. A data acquisition module is integrated into the elevator's floor call button circuit. When a button is pressed, the acquisition module records the signal from the corresponding floor, converts it into a digital signal, and transmits it to the central control system. The system also uses a clock chip to record the call time and determines the call direction based on the button type (up or down). By collecting and analyzing this data, floor call distribution data can be obtained.

[0054] S20. Based on the pre-trained AI scene recognition model, identify the type of operation scene in which the elevator is located according to the current operating parameters. The scene types include peak congestion scene, off-peak energy-saving scene, emergency braking scene, and normal stable scene;

[0055] Preferably, the AI scene recognition model is constructed by the following steps:

[0056] Collect training data sets containing different load distributions, environmental parameters, and user behavior patterns;

[0057] Data collection period: covers full-time operating data on typical weekdays (morning rush hour 7:30-9:00, lunchtime 12:00-13:30, evening rush hour 17:30-19:00), weekends, and holidays; parameter collection accuracy is generally as follows: operating speed: ±0.01m / s; load: ±1% FS; temperature and humidity: temperature: ±0.5℃, humidity: ±3% RH; steel rope deformation: ±0.01mm.

[0058] After collecting data, the human body detection algorithm (based on YOLOv8) of the elevator camera is used to count the number of passengers and their entry and exit directions. Combined with the floor call time series analysis, group behavior characteristics (such as collective commuting, concentrated meeting end, etc.) are analyzed to complete the labeling of user behavior patterns.

[0059] A convolutional neural network is used to extract the spatiotemporal correlation features of elevator operating parameters, and the attention mechanism is used to strengthen the weights of key scene features.

[0060] The feature fusion strategy adopts: constructing time series features through sliding windows (such as 5 seconds / frame) in the time dimension, and fusing the load distribution in the car (multi-area weighing sensor data) and the spatial distribution of floor calls in the spatial dimension.

[0061] Use the hierarchical clustering algorithm to divide the scene type boundaries and output the scene classification probability matrix;

[0062] Extract the feature vectors output by the CNN, determine the hierarchical clustering parameter configuration, and generate scene classification labels corresponding to the four scene types. Output the scene classification probability matrix and construct a probability density model for each scene category.

[0063] When the matching degree between the real-time operation parameters and the preset scenario is lower than the threshold, the incremental learning mechanism is triggered to update the model parameters.

[0064] Model update judgment during real-time scene recognition: when more than 10% of the sample matching degrees are lower than the threshold, the incremental learning mechanism is triggered to update the model parameters.

[0065] S30. Match the corresponding AI control strategy according to the identified scene type, calculate the brake pressure of the elevator using the brake pressure control model, and calculate the response time of the elevator using the response time control model;

[0066] Develop corresponding AI control strategies for different scenario types. For example, in peak congestion scenarios, the control strategy is to prioritize responding to calls from multiple floors to reduce the time of a single stop; in off-peak energy-saving scenarios, the elevator speed is reduced to reduce energy consumption. At the same time, a brake pressure control model and a response time control model are established, and the model parameters are optimized according to different scenarios. According to the identified scenario type, the corresponding strategy is matched from the control strategy library. The brake pressure control model calculates the appropriate brake pressure based on parameters such as the elevator speed and car load through a pre-set algorithm (such as calculation based on a physical model or prediction based on machine learning); the response time control model combines information such as floor call distribution data and the current position of the elevator to calculate the optimal response time.

[0067] S40. Obtain a maximum rebound distance when the elevator stops, calculate a compensation factor based on the maximum rebound distance and the elevator's operating parameters, use the brake compensation factor to update the calculated brake pressure and response time, and use the updated brake pressure and response time to adaptively adjust the elevator's operating parameters.

[0068] A displacement sensor (such as a laser displacement sensor or a magnetostrictive displacement sensor) is installed at the bottom of the elevator car. When the elevator comes to a complete stop, the sensor monitors the car's rebound in real time, recording and capturing the maximum rebound distance. A compensation factor is calculated using a specific formula (such as a weighted average formula) based on operating parameters such as the maximum rebound distance, elevator speed, and car load. This compensation factor is used to adjust and update the previously calculated brake pressure and response time to better meet actual operating requirements. The updated brake pressure and response time are fed back to the elevator control system, which uses these parameters to adaptively adjust operating parameters such as the elevator's speed and braking system, achieving optimal operation in different scenarios.

[0069] Therefore, by real-time monitoring of parameters such as the dynamic deformation of the steel rope, potential safety hazards can be promptly identified. In emergency braking scenarios, the system accurately calculates brake pressure and response time, and makes compensation adjustments based on rebound conditions, effectively preventing accidents such as elevator slippage and roof collisions. In peak congestion scenarios, the system optimizes elevator dispatch strategies to quickly respond to calls from multiple floors and reduce passenger wait times. Intelligently adjust operating parameters based on the scenario to make elevator operation more efficient and reasonable. In off-peak energy-saving scenarios, the system reduces elevator speed and energy consumption, reducing unnecessary energy consumption, aligning with green energy conservation concepts and lowering operating costs.

[0070] In one embodiment, the brake pressure control model building process includes:

[0071] Calculate a first relationship function of the influence of speed and acceleration on brake pressure based on the elevator acceleration, the influence weight of speed on brake pressure and the influence weight of brake pressure;

[0072] Calculate a second relationship function of the effects of temperature and humidity on brake pressure based on the reference temperature and humidity and the actual temperature and humidity collected in real time during elevator operation;

[0073] ;

[0074] ;

[0075] Where, 、 are the first relation function and the second relation function respectively, For speed The weight of the impact on the brake pressure, is the acceleration Impact weight on brake pressure; is a natural constant; are the reference temperature and reference humidity, 、 The temperature and humidity collected in real time during elevator operation are respectively are the temperature change sensitivity coefficient and the humidity change sensitivity coefficient, respectively. .

[0076] Obtain the static load and car mass of the elevator, and construct a brake pressure control model based on the static load, car mass, the first relationship function, and the second relationship function. Specifically,

[0077] ;

[0078] Where, is the brake pressure, is the adjustment coefficient, ; is the static load of the elevator, is the mass of the elevator car, is the acceleration of the elevator; is the relationship function between the speed and acceleration on the brake pressure, that is, the first relationship function, is the relationship function of the influence of temperature and humidity on brake pressure, that is, the second relationship function.

[0079] In the above model, This is an adjustment factor used to ensure sufficient braking force in extreme situations (such as emergency braking). This factor allows the system to adjust the brake pressure according to actual conditions to meet safety requirements. It is the static load of the elevator, that is, the weight of the elevator when it is empty plus the total weight of passengers or cargo. It represents the dynamic load increment caused by the acceleration of the elevator. This represents the total weight of the elevator at rest plus the inertial force due to acceleration. This reflects the total load actually carried by the elevator and is one of the fundamental factors in determining the required braking force. The required braking force increases with increasing load or acceleration / deceleration.

[0080] exist In the calculation model of The part represents the effect of speed on brake pressure, where is the influence weight of speed. As speed increases, the exponential decay factor decreases, which means that greater braking force is required at high speeds. The effect of acceleration is emphasized, especially the absolute value sign indicating that both acceleration and deceleration affect the brake pressure. The natural logarithm function is introduced here to smooth the nonlinear growth under high acceleration. are the temperature change sensitivity coefficient and the humidity change sensitivity coefficient, respectively. , this value is small, indicating that environmental factors have an impact but are not the main factor.

[0081] By integrating all of the aforementioned factors, the brake pressure regulation model attempts to capture the various dynamic and static conditions that affect brake performance. First, a base braking force is determined based on the elevator's fundamental physical characteristics, such as load, mass, and motion state. Then, changes in speed and acceleration are considered, as these parameters directly affect the energy consumption and friction requirements during braking. Finally, external environmental conditions, such as temperature and humidity, are considered, as these conditions may alter the physical properties of the brake material and, in turn, affect the braking effect.

[0082] This model enables precise prediction and control of brake pressure to adapt to varying operating conditions, ensuring safe and reliable braking in all possible scenarios. This not only improves the safety of the elevator system but also enhances its ability to respond to complex operating conditions.

[0083] In one embodiment, the process of constructing the response time control model includes:

[0084] Calculate the temperature difference between the reference temperature and the actual temperature, and the humidity difference between the reference humidity and the actual humidity respectively;

[0085] The maximum allowable response time of the elevator operation is obtained, and a response time control model is constructed based on the maximum allowable response time, acceleration, speed, temperature difference, and humidity difference of the elevator operation.

[0086] Specifically, the following formula is used to achieve this:

[0087] ;

[0088] Where, is the response time, is the maximum allowed response time, is a constant, .

[0089] Acceleration is one of the key parameters that influences the need for braking. Greater acceleration means faster response time is required to ensure a safe stop. Accurately reflects the response time requirements under different acceleration conditions. and humidity It has a significant impact on the material properties and the working state of the mechanical system. For example, high temperature may cause the material to soften or the lubrication effect to deteriorate, while high humidity may cause corrosion or increase friction. By introducing the temperature change sensitivity coefficient and humidity change sensitivity coefficient , and reference values , which can dynamically adjust the response time to adapt to different environmental conditions.

[0090] Velocity squared term Reflects the effect of kinetic energy. Higher speed means greater kinetic energy, so a shorter response time is required to ensure timely braking. To prevent the denominator from being zero and to ensure that the response time is always positive, a very small constant is introduced into the model This improves the stability of the model and ensures a reasonable response time in all situations.

[0091] Therefore, the aforementioned response time control model accurately predicts the required braking response time, ensuring timely brake response in all circumstances and preventing safety accidents caused by slow response. This is particularly true in emergency situations, such as those at high speeds or under heavy loads. Braking can be initiated quickly and stably, providing a smoother ride and reducing the discomfort associated with sudden braking. This also avoids excessive use or unnecessary delays in the braking system, improving overall system efficiency and enabling elevators to brake in a shorter time, reducing waiting time and energy consumption. Precisely controlling response time slows the aging of key components, reduces the incidence of failures, and indirectly reduces repair and replacement costs, extending the service life of the rope and other related components.

[0092] In one embodiment, the step of calculating the compensation factor based on the maximum rebound distance and the operating parameters of the elevator and updating the calculated brake pressure and response time using the brake compensation factor includes:

[0093] Calculate the compensation factor based on the elevator's maximum rebound distance, speed, and speed sensitivity coefficient;

[0094] ;

[0095] Where, is the compensation factor, is the maximum rebound distance, is the speed sensitivity coefficient, ;

[0096] Multiplying the calculated brake pressure by the compensation factor to obtain an updated brake pressure;

[0097] Divide the calculated response time by the compensation factor to get the updated response time, which is:

[0098] ;

[0099] ;

[0100] Where, 、 Updated brake pressure and response time, respectively.

[0101] In the above model, the speed of the elevator It has a direct impact on the braking effect. When driving at high speed, stronger braking force and faster response time are required, while the opposite is true when driving at low speed. This part can capture the impact of speed on braking demand and make appropriate adjustments. An important indicator to measure the energy release of an elevator at the moment of stopping. A larger rebound distance means that the system has not fully absorbed the kinetic energy, posing a safety hazard. The compensation factor takes into account the change of rebound distance with speed, ensuring that sufficient braking force can be provided even in the case of large rebound. It can dynamically adjust the braking pressure and response time according to actual operating conditions, making the braking process smoother and safer.

[0102] Therefore, the compensation factor can precisely adjust brake pressure and response time based on real-time data, further optimizing the control model's calculations to ensure sufficient braking force in all circumstances. This prevents safety incidents caused by insufficient or excessive braking, and provides a smoother ride, especially at high speeds or under heavy loads, reducing the discomfort caused by sudden braking. This can slow the aging of key components, reduce the incidence of failures, indirectly reduce repair and replacement costs, and extend the service life of the steel rope and other related components. Finally, the compensated brake pressure and response time can avoid excessive use or unnecessary delays in the braking system, improving overall system efficiency and achieving a safer, more comfortable, and efficient braking process.

[0103] In one embodiment, the pre-trained AI scene recognition model identifies the type of operating scene the elevator is in according to current operating parameters, including:

[0104] Determine whether the actual temperature collected in real time during elevator operation exceeds a first preset value. When the temperature exceeds the first preset value, obtain the load in the elevator car and identify the type of operation scenario of the elevator based on the load.

[0105] Based on the technical specifications and safety standards provided by the steel rope manufacturer, a reasonable temperature threshold, known as the first preset value, is determined. This threshold should be unlikely to be reached under normal operating conditions, yet sensitive enough to detect potential risks. If the steel rope's operating temperature exceeds this threshold, it indicates a potential safety hazard, necessitating rapid determination of the target landing floor for braking.

[0106] Preferably, after obtaining the load in the elevator car, the method further includes matching the target stop floor according to the load, including:

[0107] When the load is greater than a second preset value, the floor closest to the current position is used as the target docking floor;

[0108] When the load is less than or equal to a second preset value, the median floor of the called floors in the elevator car is used as the target stopping floor.

[0109] In this embodiment, when the load is greater than the second preset value, the floor closest to the current position is used as the target docking floor;

[0110] When the load is less than or equal to a second preset value, the median floor of the called floors in the elevator car is used as the target stopping floor.

[0111] Based on the elevator's design load capacity and safety standards, determine a reasonable load threshold, the second preset value. This threshold should be the maximum load acceptable within the normal operating range, ensuring that additional safety measures are taken when this value is exceeded.

[0112] When the load is greater than the second preset value, the elevator is considered to be in a heavy-loaded state; when the load is less than or equal to the second preset value, the elevator is considered to be in a light-loaded state. The following different target stop floor matching methods are provided for different loads:

[0113] Overload (load > second preset value): Once the load exceeds the second preset value, the central control system immediately locates the nearest floor and sets it as the target landing floor. To evacuate passengers as quickly as possible, the system accelerates the landing instruction, ensuring that the elevator reaches the target floor in the shortest possible time while maintaining the necessary stability to avoid causing discomfort to passengers.

[0114] Lightly loaded (load ≤ second preset value): While the elevator is in operation, it records all floor information from call button presses inside the car. The central control system calculates a median floor from the recorded call floors as the target landing floor. The median floor is the middle floor among all called floors; if there is an even number of called floors, the lower of the two middle floors is selected. Based on the selected median floor, the control command is generated to ensure the elevator moves smoothly to that floor and opens the doors. For example, if the call floors are 3, 5, 6, 7, and 9, floor 6 may be selected as the target landing floor.

[0115] In one embodiment, the method further includes adopting different execution logics for different operation scenario types, including:

[0116] In peak congestion scenarios, priority is given to shortening response time and increasing braking force, while simultaneously activating the car pre-dispatching algorithm to reduce inter-floor dwellings;

[0117] Response Time Optimization: When the AI scene recognition model identifies a peak congestion scenario, the system prioritizes the highest-priority floor call requests in the current queue. Based on the elevator's current position, direction of travel, and call distribution data for each floor, a dynamic programming algorithm is used to replan the optimal route, reducing unnecessary stops. Furthermore, the elevator control system's signal processing frequency is increased, shortening the response time calculation cycle from the typical 1 second to 0.5 seconds, speeding up responses to new call requests.

[0118] Braking force enhancement: The brake pressure control model increases braking pressure by 20%-30% based on existing calculations, based on peak scenario parameters. For example, by increasing the electromagnetic force of the traction machine brake, the elevator can stop faster and more steadily when docking, reducing waiting time for passengers entering and exiting the elevator car.

[0119] The car pre-dispatch algorithm is activated: Based on historical floor call data for the same period and current real-time call data, it predicts the floors likely to receive calls in the future. When an elevator is empty or about to be empty, it pre-dispatchs it to floors with predicted high call volumes. For example, during the weekday morning rush hour, if high call demand is predicted on the upper floors of an office building, idle cars can be pre-dispatched to the mid- and upper floors to be on standby, reducing travel time between floors for subsequent calls.

[0120] By prioritizing high-priority calls, optimizing route planning, and pre-dispatching elevators, average passenger waiting time and elevator cycle times are significantly reduced, significantly improving peak-hour capacity and alleviating congestion. Increased braking force ensures more stable elevator stops, reducing safety risks such as falls caused by rapid starts and stops or crowded passengers, and ensuring passenger safety. Rapid response and efficient operation reduce waiting times in elevator halls and elevator cars, enhancing the riding experience and increasing passenger satisfaction.

[0121] In low-peak energy-saving scenarios, a segmented soft braking strategy is adopted, combined with coasting to reduce motor energy consumption;

[0122] Implementation of a segmented soft braking strategy: When the elevator is in an off-peak energy-saving scenario, the braking system divides the braking process into multiple stages based on the elevator's current speed and distance to the target floor. When the elevator is farther from the target floor, a smaller braking force is used to slowly decelerate. As the elevator approaches the target floor, the braking force is gradually increased until it reaches a smooth stop. For example, when the elevator is running at rated speed and three floors away from the target floor, it will first decelerate at 30% of the normal braking force. When the elevator is one floor away from the target floor, the braking force is increased to the normal level for a smooth stop.

[0123] Coasting: Leveraging the elevator's inertia, the system preemptively shuts down the motor drive, ensuring safety and comfort, allowing the elevator to coast to the target floor before braking. The system monitors the elevator's speed, load, and distance to the target floor in real time to calculate the optimal motor shutdown timing and coasting distance. For example, if the elevator is unloaded, close to the target floor, and moving at a low speed, the motor is shut down 1-2 seconds in advance, allowing the elevator to coast to a stop, reducing motor operation time.

[0124] Segmented soft braking reduces energy loss during braking, while coasting reduces motor operating time. Combined, these two features significantly reduce energy consumption during off-peak hours, lowering operating costs and aligning with the concept of green energy conservation. Soft braking and coasting reduce frequent starts and stops, as well as the intense stresses on components like the motor and brake, reducing component wear, extending the life of the elevator equipment, and reducing maintenance costs and frequency. The smooth segmented braking and coasting process avoids the jerking sensation caused by sudden stops and starts, providing passengers with a more comfortable ride.

[0125] In emergency braking scenarios, redundant braking units are triggered, and the braking trajectory is simulated through the digital twin system to avoid resonance risks.

[0126] Redundant braking unit triggering: Once the AI scenario recognition model detects an emergency braking scenario (such as an elevator overspeed or rope breakage), it immediately triggers the elevator's redundant braking unit. This unit typically includes a backup brake and safety caliper. If the primary braking system fails or in an emergency, it activates quickly, mechanically forcing the elevator to a stop.

[0127] The digital twin system simulates braking trajectories: Simultaneously, the digital twin system rapidly constructs a virtual model of the elevator's current state based on the elevator's real-time operating parameters (such as speed, load, and rope status) and the physical model. This virtual model is used to simulate braking trajectories under different braking strategies and predict potential resonance risk points. For example, by varying the braking pressure application method and speed, the system simulates the braking process, analyzes the vibration of components such as the car and rope, and identifies the optimal braking solution that avoids resonance. This solution is then fed back to the elevator control system to guide actual braking operations.

[0128] The rapid activation of redundant braking units provides additional safety in emergency situations. Even if the main braking system fails, the elevator can be brought to a rapid stop, preventing serious accidents and protecting passengers. By simulating the braking trajectory through the digital twin system, resonance risks can be proactively avoided, preventing further damage to elevator components or secondary accidents caused by resonance, thereby reducing equipment damage and economic losses. This mechanism enables intelligent emergency braking, eliminating the need for excessive human intervention from brake initiation to solution optimization. This improves the efficiency and accuracy of elevator emergency response and provides data support and decision-making basis for subsequent rescue and troubleshooting.

[0129] See also Figure 2 In one embodiment, the present invention further provides a novel elevator multi-scenario operation intelligent control system based on an AI model, the system comprising:

[0130] The data acquisition unit 100 is used to collect the operating parameters of the elevator in real time, including the elevator speed, car load, ambient temperature and humidity, dynamic deformation of the steel rope, and floor call distribution data;

[0131] The scene recognition unit 200 is used to identify the type of operation scene of the elevator based on the pre-trained AI scene recognition model and the current operation parameters. The scene types include peak congestion scene, off-peak energy-saving scene, emergency braking scene and normal stable scene;

[0132] The strategy matching unit 300 is used to match the corresponding AI control strategy according to the identified scenario type, calculate the brake pressure of the elevator using the brake pressure control model, and calculate the response time of the elevator using the response time control model;

[0133] The parameter adjustment unit 400 is used to obtain the maximum rebound distance when the elevator stops, calculate the compensation factor based on the maximum rebound distance and the operating parameters of the elevator, use the brake compensation factor to update the calculated brake pressure and response time, and use the updated brake pressure and response time to adaptively adjust the operating parameters of the elevator.

[0134] It can be understood that the functions or modules included in the system provided in this embodiment can be used to execute the method described in the above method embodiment. Its specific implementation can refer to the description of the above method embodiment. For the sake of brevity, it will not be repeated here.

[0135] The present invention also provides an electronic device, including a processor and a memory, wherein the memory is used to store computer program code, and the computer program code includes computer instructions. When the processor executes the computer instructions, the electronic device executes a method as described in any one of the possible implementation modes.

[0136] The present invention also provides a computer-readable storage medium, in which a computer program is stored. The computer program includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes a method as described in any one of the possible implementation methods described above.

[0137] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

Claims

1. A novel intelligent control method for multi-scenario operation of elevators based on AI model, characterized in that: The method comprises: Real-time collection of elevator operating parameters, including elevator speed, car load, ambient temperature and humidity, dynamic deformation of steel ropes, and floor call distribution data; Based on the pre-trained AI scene recognition model, the elevator's operating scenario type is identified according to the current operating parameters. The scenario types include peak congestion scenarios, off-peak energy-saving scenarios, emergency braking scenarios, and normal stable scenarios. The AI control strategy is matched to the identified scenario type, the brake pressure of the elevator is calculated using the brake pressure control model, and the response time of the elevator is calculated using the response time control model. The maximum rebound distance of the elevator when it stops is obtained, and a compensation factor is calculated based on the maximum rebound distance and the elevator's operating parameters. The calculated brake pressure and response time are updated using the brake compensation factor, and the updated brake pressure and response time are used to adaptively adjust the elevator's operating parameters.

2. The novel elevator multi-scenario operation intelligent control method based on AI model according to claim 1 is characterized in that: The brake pressure control model construction process includes: Calculate a first relationship function of the influence of speed and acceleration on brake pressure based on the acceleration of the elevator and the influence weight of the speed on brake pressure; Calculate a second relationship function of the effects of temperature and humidity on brake pressure based on the reference temperature and reference humidity and the actual temperature and humidity collected in real time during elevator operation; The static load and car mass of the elevator are obtained, and a brake pressure control model is constructed according to the static load, car mass, the first relationship function and the second relationship function.

3. The novel elevator multi-scenario operation intelligent control method based on AI model according to claim 2 is characterized in that: The process of constructing the response time control model includes: Calculate the temperature difference between the reference temperature and the actual temperature, and the humidity difference between the reference humidity and the actual humidity respectively; The maximum allowable response time of the elevator operation is obtained, and a response time control model is constructed based on the maximum allowable response time, acceleration, speed, temperature difference, and humidity difference of the elevator operation.

4. The novel elevator multi-scenario operation intelligent control method based on AI model according to claim 3 is characterized in that: The method of calculating the compensation factor according to the maximum rebound distance and the operating parameters of the elevator and updating the calculated brake pressure and response time using the brake compensation factor includes: Calculate the compensation factor based on the elevator's maximum rebound distance, speed, and speed sensitivity coefficient; Multiplying the calculated brake pressure by the compensation factor to obtain an updated brake pressure; Divide the calculated response time by the compensation factor to obtain the updated response time.

5. The novel elevator multi-scenario operation intelligent control method based on AI model according to claim 2 is characterized in that: The pre-trained AI scene recognition model identifies the type of operating scene the elevator is in based on current operating parameters, including: Determine whether the actual temperature collected in real time during elevator operation exceeds a first preset value. When the temperature exceeds the first preset value, obtain the load in the elevator car and identify the type of operation scenario of the elevator based on the load.

6. The novel elevator multi-scenario operation intelligent control method based on AI model according to claim 5 is characterized in that: After obtaining the load in the elevator car, the target stop floor is matched according to the load, including: When the load is greater than a second preset value, the floor closest to the current position is used as the target docking floor; When the load is less than or equal to a second preset value, the median floor of the called floors in the elevator car is used as the target stopping floor.

7. The novel elevator multi-scenario operation intelligent control method based on AI model according to claim 1 is characterized in that: The method further includes adopting different execution logics for different operation scenario types, including: In peak congestion scenarios, priority is given to shortening response time and increasing braking force, while simultaneously activating the car pre-dispatching algorithm to reduce inter-floor dwellings; In low-peak energy-saving scenarios, a segmented soft braking strategy is adopted, combined with coasting to reduce motor energy consumption; In emergency braking scenarios, redundant braking units are triggered, and the braking trajectory is simulated through the digital twin system to avoid resonance risks.

8. A new type of elevator multi-scenario operation intelligent control system based on AI model, characterized by: The system comprises: Data acquisition unit, used to collect real-time operating parameters of the elevator, including elevator speed, car load, ambient temperature and humidity, dynamic deformation of the steel rope, and floor call distribution data; A scene recognition unit is used to identify the type of operating scene the elevator is in based on the pre-trained AI scene recognition model and the current operating parameters. The scene types include peak congestion scenes, off-peak energy-saving scenes, emergency braking scenes, and normal stable scenes; The strategy matching unit is used to match the corresponding AI control strategy according to the identified scenario type, calculate the elevator's brake pressure through the brake pressure control model, and calculate the elevator's response time based on the response time control model; The parameter adjustment unit is used to obtain the maximum rebound distance when the elevator stops, calculate the compensation factor based on the maximum rebound distance and the elevator's operating parameters, use the brake compensation factor to update the calculated brake pressure and response time, and use the updated brake pressure and response time to adaptively adjust the elevator's operating parameters.

9. An electronic device, characterized in that: include: A processor and a memory, the memory being used to store computer program code, the computer program code including computer instructions, and when the processor executes the computer instructions, the electronic device executes the new elevator multi-scenario operation intelligent control method based on the AI model as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes the new elevator multi-scenario operation intelligent control method based on the AI model as described in any one of claims 1 to 7.

Images