Tray stress monitoring and early warning method and device of automobile elevator and storage medium

Abstract

The invention relates to the technical field of stress monitoring, and discloses a tray stress monitoring and early warning method and device of an automobile elevator and a storage medium. The method comprises the steps that intelligent trays are installed at the four tail ends of a lifting machine supporting arm respectively, an intelligent internet-of-things control box is deployed, and wireless communication connection between the intelligent internet-of-things control box and the intelligent trays is established; setting tray stress judgment logic of the intelligent tray and constructing an initial multi-dimensional judgment model; the first stress data transmitted by the four intelligent trays are collected and subjected to stress data analysis, and a comprehensive analysis result is obtained; carrying out early warning level judgment and executing safety control measures; and recording and analyzing the second stress data of each intelligent tray to obtain a dynamically optimized early warning threshold and lifting process parameters, and updating the initial multi-dimensional judgment model to obtain a target multi-dimensional judgment model, thereby improving the accuracy of tray stress monitoring and early warning of the automobile elevator.

Description

Technical Field

[0001] The present application relates to the technical field of force monitoring, and in particular to a method, device and storage medium for monitoring and warning the force of a tray of a car lift. Background Art

[0002] The lifting structure of a two-post lift consists of four arms and a total of four trays in contact with the vehicle body. However, in actual use, it is often the case that only three of them are under stress, while the other one is suspended in the air or bounced off, causing safety hazards and serious vehicle crashes, which are not uncommon.

[0003] Currently, the existing technology lacks a real-time monitoring and early warning system, and is unable to promptly detect and handle abnormal pallet stress. Maintenance personnel often rely on experience to determine whether the lifting process is safe. This subjective judgment method has great uncertainty and risk, which leads to low accuracy of the existing technology. Summary of the invention

[0004] The present application provides a method, device, equipment and storage medium for monitoring and warning the force of a pallet of a car lift, which are used to improve the accuracy of monitoring and warning the force of a pallet of a car lift.

[0005] In a first aspect, the present application provides a method for monitoring and warning the force on a tray of a car lift, the method comprising: Smart pallets with pressure sensing and wireless communication functions are installed at the four ends of the lift support arm respectively. At the same time, a smart IoT control box is deployed in the lift control area to establish a wireless communication connection between the smart IoT control box and the smart pallet; The smart IoT control box sets the pallet force judgment logic of the smart pallet and constructs an initial multi-dimensional judgment model including a minimum force threshold, a force balance degree, and a force change rate; The intelligent IoT control box collects the first force data transmitted by the four intelligent pallets, and analyzes the force data through the initial multi-dimensional judgment model to obtain a comprehensive analysis result, which includes: the lifting machine movement state and the abnormal force state information of each intelligent pallet; The intelligent IoT control box determines the warning level and executes corresponding safety control measures according to the comprehensive analysis results, and the safety control measures include issuing a warning display, triggering an alarm, and sending a lift control instruction; After executing the corresponding safety control measures, the intelligent IoT control box records and analyzes the second force data of each smart pallet to obtain dynamically optimized warning thresholds and lifting process parameters, and updates the initial multi-dimensional judgment model through the dynamically optimized warning thresholds and lifting process parameters to obtain the target multi-dimensional judgment model.

[0006] In a second aspect, the present application provides a device for monitoring and warning the force on a pallet of a car lift, the device comprising: An installation module is used to install smart pallets with pressure sensing and wireless communication functions at the four ends of the lift support arm, and at the same time, a smart IoT control box is deployed in the lift control area to establish a wireless communication connection between the smart IoT control box and the smart pallet; A construction module, used for the smart IoT control box to set the pallet force judgment logic of the smart pallet and to construct an initial multi-dimensional judgment model including a minimum force threshold, a force balance degree and a force change rate; An analysis module, used for the smart IoT control box to collect the first force data transmitted by the four smart pallets, and to analyze the force data through the initial multi-dimensional judgment model to obtain a comprehensive analysis result, wherein the comprehensive analysis result includes: the lifting machine motion state and the abnormal force state information of each smart pallet; An early warning module, used for the intelligent IoT control box to determine the early warning level and execute corresponding safety control measures according to the comprehensive analysis results, and the safety control measures include issuing a warning display, triggering an alarm and sending a lift control instruction; An updating module is used for recording and analyzing the second force data of each smart pallet by the smart IoT control box after executing corresponding safety control measures, obtaining dynamically optimized early warning thresholds and lifting process parameters, and updating the initial multi-dimensional judgment model through the dynamically optimized early warning thresholds and lifting process parameters to obtain a target multi-dimensional judgment model.

[0007] The third aspect of the present application provides a pallet force monitoring and early warning device for a car lift, comprising: a lift column, a lift support arm connected to the lift column; four smart pallets, respectively installed at the four ends of the lift support arm, each smart pallet having pressure sensing and wireless communication functions; an intelligent Internet of Things control box, arranged in the lift control area, the intelligent Internet of Things control box establishing a wireless communication connection with the four smart pallets, so that the pallet force monitoring and early warning device for the car lift executes the above-mentioned pallet force monitoring and early warning method for the car lift.

[0008] A fourth aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned method for monitoring and warning the force on the tray of a car lift.

[0009] In the technical solution provided by the present application, by installing an intelligent tray at the end of the lift arm and combining it with an intelligent IoT control box, the real-time collection and transmission of the force data of the four trays is realized, and a multi-dimensional judgment model including the minimum force threshold, force balance and force change rate is adopted to comprehensively evaluate the force condition of the tray, thereby improving the accuracy and reliability of abnormal detection. The force data is processed to accurately identify the motion state of the lift and the force abnormality of each tray. According to the degree of abnormality, the judgment of three levels of abnormal state, namely, slight, moderate and severe, is realized, and corresponding early warning and control strategies are adopted for different levels, effectively reducing the safety risk. By recording and analyzing the second force data, the system can dynamically optimize the early warning threshold and lifting process parameters, update the judgment model, and optimize the lifting process parameter model. By analyzing and evaluating each intelligent tray separately, the system can adapt to the needs of different models and different usage scenarios, effectively reducing the impact of data noise, improving the accuracy of abnormal detection, building a complete safety monitoring chain, and comprehensively improving the accuracy of the early warning of the force monitoring of the tray of the car lift. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0011] Figure 1 A schematic diagram of an embodiment of a method for monitoring and warning the force on a tray of a car lift in an embodiment of the present application; Figure 2 It is a schematic diagram of an embodiment of a device for monitoring and warning the force on a tray of a car lift in an embodiment of the present application; Figure 3 It is a schematic diagram of an embodiment of a force monitoring and early warning device for a tray of a car lift in an embodiment of the present application. DETAILED DESCRIPTION

[0012] The embodiments of the present application provide a method, device, equipment and storage medium for monitoring and warning the force of a tray of a car lift. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0013] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiment of the present application, an embodiment of the method for monitoring and warning the force on the tray of the car lift includes: Step S101: smart pallets with pressure sensing and wireless communication functions are installed at the four ends of the lift support arm respectively. At the same time, a smart IoT control box is deployed in the lift control area to establish a wireless communication connection between the smart IoT control box and the smart pallet; It is understandable that the execution subject of the present application may be a vehicle lift tray force monitoring and early warning device, or a terminal or a server, which is not specifically limited here. The present application embodiment is described by taking a server as the execution subject as an example.

[0014] Specifically, smart pallets with pressure sensing and wireless communication functions are installed at the four ends of the lift arm to monitor the pressure on the pallets in real time. These smart pallets need to integrate wireless communication modules to ensure that pressure data can be transmitted to the control center. The design of the pallets is optimized so that they can be firmly fixed at the ends of the lift arm to ensure that they will not be displaced or fall off during the lifting process. An intelligent IoT control box is deployed in the control area of ​​the lift. The control box should have data processing capabilities and wireless communication functions to receive and process pressure data from the smart pallets. During the installation process, the control box needs to be connected to the power system and control system of the lift to ensure stable operation and data communication. Signal strength and interference tests are performed to ensure that the smart pallets can maintain good communication effects at different positions and heights of the lift. During the installation and configuration of the control box, initial settings are performed through software, including device pairing, communication protocol configuration, and data acquisition frequency setting. The smart pallet and the control box need to be paired to ensure that they can recognize each other and establish a stable communication connection. Through the management software of the smart IoT control box, the status of each smart pallet is monitored in real time, and the pressure data is analyzed and processed.

[0015] Step S102: The smart IoT control box sets the tray force judgment logic of the smart tray and constructs an initial multi-dimensional judgment model including a minimum force threshold, a force balance degree, and a force change rate;

[0016] Specifically, the initial minimum force threshold of the smart pallet is set. The initial minimum force threshold is determined by analyzing the structure and material properties of the smart pallet and combining the possible minimum load requirements in practical applications. The historical force data of the four smart pallets are used to calculate the force balance threshold to obtain the initial force balance threshold. The force balance threshold is obtained by comparing the force differences of each pallet under different load conditions. A large amount of historical data is collected and stored, and statistical analysis and modeling are performed to ensure the accuracy and reliability of the calculation results. The sampling frequency of the smart pallet is set to ensure that the smart pallet can collect force data at an appropriate frequency under different working conditions, thereby ensuring the real-time and accuracy of the data. According to the sampling frequency and historical force data, the force change rate threshold is calculated to obtain the initial force change rate threshold. The force change rate threshold reflects the rate of force change of the pallet. By analyzing historical data, a reasonable change rate range is determined to avoid false alarms or missed alarms due to too fast or too slow force changes. The conditions for judging the motion state of the lift are set to obtain the judgment criteria for the three states of rising, descending and stationary. These standards need to be determined by combining the physical characteristics and operation mode of the lift, through sensor data and motion control logic, so as to accurately identify the current working state of the lift. According to the initial minimum force threshold, the initial force balance threshold and the initial force change rate threshold, the abnormal state classification standard is formulated to obtain the judgment conditions of three levels of abnormal states: slight, moderate and severe. Each level of abnormal state corresponds to different force conditions and safety risks. By grading the abnormal state, early warning and control can be carried out more accurately. At the same time, the corresponding relationship between abnormal state and early warning response is established to obtain early warning and control strategies for abnormal states of different levels. These strategies include issuing warning displays, triggering alarms and sending lift control instructions to ensure that corresponding measures can be taken in time when abnormalities are found to ensure the safety of the lift and personnel. According to the historical force data, the effective load range of the smart pallet is calculated to obtain the normal working load range of each smart pallet. The normal working load range refers to the load range that the smart pallet can withstand under normal use. This range is determined by analyzing historical data. According to the judgment criteria, judgment conditions, normal working load range and early warning and control strategies for abnormal states of different levels, the original judgment model is constructed. The weight distribution of the original judgment model is set to obtain the initial weight value of each judgment indicator in the comprehensive evaluation. The weight value is set according to the importance of each judgment indicator and the degree of influence on the overall judgment result. By reasonably allocating weights, the accuracy and reliability of the judgment model can be improved. According to the initial weight value, the parameters of the original judgment model are adjusted to obtain the initial multi-dimensional judgment model.

[0017] Step S103: the intelligent IoT control box collects the first force data transmitted by the four intelligent pallets, and analyzes the force data through an initial multi-dimensional judgment model to obtain a comprehensive analysis result, which includes: the lifting machine motion state and the abnormal force state information of each intelligent pallet; Specifically, the intelligent IoT control box collects the first force data of the four intelligent pallets respectively to obtain four sets of original force data sets. Each set of original force data sets is filtered to eliminate noise and interference, and obtain four sets of filtered force data sets that are more accurate and clear. Time series analysis is performed on each set of filtered force data sets to obtain the force change trend of each intelligent pallet. By analyzing the force change trend, the force condition of each pallet in different time periods is understood, and according to the force change trend of the four intelligent pallets, the motion state of the lift is comprehensively judged to determine the current motion state of the lift. Based on the four sets of filtered force data sets, the force balance of each intelligent pallet is calculated respectively. The force balance reflects the uniformity of force on each pallet during the lifting process, which is one of the important indicators for judging the force state of the pallet. Differential operations are performed on each set of filtered force data sets to obtain the force change rate of each intelligent pallet. The force change rate can reflect the dynamic change of the force on the pallet and is an important parameter for judging the force state of the pallet. According to the initial minimum force threshold, initial force balance threshold and initial force change rate threshold in the initial multi-dimensional judgment model, the force value, force balance and force change rate of each smart pallet are compared respectively to obtain the abnormality score of each smart pallet corresponding to each indicator. These scores reflect the abnormality of each smart pallet in different dimensions. Based on the weights in the initial multi-dimensional judgment model, the abnormality scores of each smart pallet are weighted and summed to obtain the comprehensive abnormality score of each smart pallet. The comprehensive abnormality score is a comprehensive indicator that reflects the comprehensive abnormality of each smart pallet in all dimensions. According to the comprehensive abnormality score of each smart pallet and the abnormal state grading standard in the initial multi-dimensional judgment model, the abnormal force state of each smart pallet is judged respectively to obtain the abnormal state level of each smart pallet. The lift motion state and the abnormal state level of each smart pallet are combined to obtain a comprehensive analysis result.

[0018] Perform a 5-point moving average process on each set of filtered force data sets, and smooth the data by calculating the average of every five consecutive data points to reduce noise and fluctuations, and obtain a smoother force data series. Perform a first-order difference operation on the smoothed force data series to obtain a force change series. Through the first-order difference, the trend and rate of data change are captured. Based on the least squares method, a linear regression is performed on the force change series to obtain the linear equation of the force change trend of each smart pallet. Through the least squares method, an optimal straight line is fitted, which reflects the overall trend of force changes. According to the slope of the linear equation of the force change trend , set the threshold and ,in , and classify the force change trend of each smart pallet: when the slope When , it is judged as an upward trend; when the slope When , it is judged as a downward trend; when When , it is judged as a stationary trend; when Then, the trend judgment results of the four smart pallets were statistically analyzed to obtain the frequency of occurrence of rising, falling, stationary and uncertain trends. , and . Set the weight coefficient according to the frequency of trend occurrence , and , and calculate the comprehensive trend indicator CT, the calculation formula is: ; Setting thresholds for the composite trend indicator and , and compare the threshold of the comprehensive trend indicator CT: when CT>CTu, the preliminary judgment result is an upward state; when CT When , the preliminary judgment result is a descending state; when When , the preliminary judgment result is a static state. The Kalman filter algorithm is used to perform state estimation on the preliminary judgment result to obtain a corrected motion state judgment result.

[0019] The Kalman filter algorithm is a recursive algorithm that improves the accuracy and stability of state judgment through continuous updating and correction. Based on the corrected motion state judgment results, the state sequence of the past N time points is counted, the state transition frequency is calculated, and the state transition probability matrix is ​​obtained. The state transition probability matrix reflects the transition rules between different states. By analyzing these rules, future state changes are predicted. Using the state transition probability matrix and the current state, the motion state of the lift at the next moment is predicted to obtain the current motion state of the lift. This process combines historical data and current state. Through probability statistics and prediction models, the motion state at the next moment can be accurately predicted, thereby providing a basis for system control and early warning.

[0020] The four groups of filtered force data sets are divided into time windows. By setting a fixed time interval, the continuous force data are divided into force data subsets within multiple time windows. Each time window contains a certain number of force data points, which facilitates local statistical analysis of the force data. Statistical analysis is performed on the force data subsets within each time window to calculate the average force value of each smart pallet in each time window. The average force value can reflect the force condition of the smart pallet within the time window. According to the average force value of each smart pallet in each time window, the force values ​​of the four smart pallets are sorted to obtain the sorting result from large to small, and the ratio of the maximum force value to the minimum force value in the sorting result is calculated to obtain the initial force balance degree of each time window. The initial force balance degree is screened according to the preset abnormal value range, and the abnormal values ​​beyond the preset range are eliminated, and the abnormal fluctuations or noise in the data are eliminated to obtain a more accurate force balance degree. After the screening is completed, the corrected force balance sequence is obtained. In order to further smooth the data fluctuations, the corrected force balance sequence is subjected to exponential smoothing to obtain a smoothed force balance sequence. Exponential smoothing can effectively reduce data fluctuations and make the force balance sequence more stable and continuous. Based on the smoothed force balance sequence, the change trend of the force balance is analyzed to obtain the trend index of the force balance. The trend analysis of the force balance can reveal the overall trend of the force change. According to the trend index of the force balance, the force balance is predicted to obtain the predicted force balance at the next moment. The prediction process is based on historical data and trend analysis, and a certain algorithm model is used to predict the future changes in force balance. After obtaining the current force balance and the predicted force balance, the two are weighted averaged to obtain a comprehensive force balance index. The weighted average can combine the current force balance and the future predicted force balance to make the comprehensive force balance more accurate and reliable. The comprehensive force balance index is normalized to make the force balance of different smart pallets comparable, and the final force balance of each smart pallet is obtained.

[0021] Each group of filtered force data sets is sampled at equal intervals to obtain a discrete force data sequence with a fixed time interval to ensure the uniformity and comparability of the data. The discrete force data sequence is subjected to a first-order difference operation to calculate the force difference sequence between adjacent time points. According to the sampling time interval and the force difference sequence, the instantaneous force change rate at each time point is calculated to obtain the original force change rate sequence. Since the original force change rate sequence may contain noise and abnormal fluctuations, it is subjected to median filtering to obtain a smoothed force change rate sequence. Median filtering is a commonly used signal processing technique that can effectively remove noise while retaining the key features of the signal. The smoothed force change rate sequence is segmented and the entire sequence is divided into multiple time windows, each of which contains a certain number of force change rate data points. In this way, the force changes in different time periods are analyzed independently. In each time window, the force change rate subsequence is statistically analyzed to calculate the force change rate characteristics of each time window. These characteristics include the average change rate, the standard deviation of the change rate, etc., which can reflect the overall situation of the force change in the time window. Based on the force change rate characteristics of each time window, the linear regression method is used to fit the force change rate trend line to obtain the change trend of the force change rate. Linear regression can find the most suitable straight line to reflect the overall trend of force change. According to the change trend of the force change rate, prediction is made to obtain the predicted force change rate sequence. By analyzing historical data and current trends, future force changes are predicted. The smoothed force change rate sequence is weightedly fused with the predicted force change rate sequence to obtain a comprehensive force change rate sequence. Combine current data and predicted data to improve the accuracy and reliability of the force change rate. The comprehensive force change rate sequence is normalized to make the force change rates of different smart pallets comparable, and the final force change rate of each smart pallet is obtained.

[0022] Step S104: the intelligent IoT control box determines the warning level and executes corresponding safety control measures according to the comprehensive analysis results. The safety control measures include issuing a warning display, triggering an alarm, and sending a lift control command.

[0023] Specifically, the smart IoT control box receives and processes the force data from the smart pallet in real time, including the current motion state of the lift and the abnormal force state information of each smart pallet. The smart IoT control box determines the warning level of the abnormal force state information. The warning level is determined based on preset standards and thresholds, such as the magnitude of the force change rate, the abnormal degree of force balance, and the change trend of the comprehensive force data. By comparing the current data with these thresholds, the smart IoT control box can determine which level of warning the current situation belongs to. The warning level is usually divided into three categories: mild, moderate and severe, and each level corresponds to a different degree of risk and treatment measures. At the mild warning level, the smart IoT control box issues a warning display. This can be achieved through the display on the control box, the connected computer monitor, or the application interface on the mobile device. The warning display content includes the current force data, the description of the abnormal situation, and the recommended operating instructions, reminding the operator to pay attention but no immediate action is required. When the warning level rises to moderate, the smart IoT control box not only issues a warning display, but also triggers an alarm. The alarm system can be an audible and visual alarm that draws the operator's attention by emitting harsh sounds and flashing lights, ensuring a response in a short period of time. The purpose of the alarm is to alert the operator to potential risks that need to be checked and handled as soon as possible. If the warning level reaches severe, the smart IoT control box sends a lift control command to directly control the operation status of the lift through the built-in control module. The control command can include stopping the lift, reducing the lift height, or returning the lift to its original position to prevent further danger.

[0024] Step S105: After executing the corresponding safety control measures, the smart IoT control box records and analyzes the second force data of each smart pallet to obtain dynamically optimized warning thresholds and lifting process parameters, and updates the initial multi-dimensional judgment model through the dynamically optimized warning thresholds and lifting process parameters to obtain the target multi-dimensional judgment model.

[0025] Specifically, the second force data of each smart pallet is segmented into time windows, and the continuous force data is divided into multiple time windows of fixed length for local statistical analysis. Each time window contains several data points. In this way, the analysis of the force data is refined. A statistical analysis is performed on the subset of force data in each time window, and the statistical characteristics such as the mean value, standard deviation, maximum value, and minimum value of each window are calculated to reflect the force situation in each time window. According to the statistical characteristics of each window, pattern recognition is performed on the second force data of each smart pallet. The pattern recognition method can adopt cluster analysis, principal component analysis and other technologies. By classifying the force data under different working conditions, the force mode of each smart pallet under different working conditions is identified, and the typical characteristics and laws of the force under various working conditions are summarized. The force mode of each smart pallet under different working conditions is compared and analyzed, and the differences in force characteristics under various working conditions are identified. Based on this analysis, the optimized abnormal judgment standard is obtained. The optimized abnormal judgment standard is more suitable for actual working conditions and can more accurately identify abnormal situations. Based on the optimized abnormal judgment criteria, the initial minimum force threshold, initial force balance threshold and initial force change rate threshold in the initial multi-dimensional judgment model are dynamically adjusted to obtain new warning thresholds, which include the target minimum force threshold, target force balance threshold and target force change rate threshold. The historical force data and the corresponding lift operation status are trained, and a model is trained using machine learning algorithms (such as decision trees, support vector machines, neural networks, etc.) in combination with the dynamically optimized warning threshold to obtain the lifting process parameters. The model can predict the force state and potential risks in the lifting process under given warning threshold conditions. According to the dynamically optimized warning threshold and lifting process parameters, the judgment logic of the initial multi-dimensional judgment model is reconstructed, and the judgment logic is updated so that it can make more accurate judgments based on the latest warning threshold and process parameters. The reconstruction process includes redesigning the judgment rules and conditions to ensure that the judgment model can operate correctly under the new threshold conditions. Subsequently, the weights of each indicator in the updated judgment logic are optimized, and the weight distribution scheme of each indicator is determined using optimization algorithms (such as weighted least squares method, genetic algorithm, etc.) to ensure that the importance of each indicator in the comprehensive judgment is reasonably reflected. Through weight optimization, the overall accuracy and stability of the judgment model are improved. Based on the new weight distribution scheme, the parameters of the updated judgment logic are fine-tuned to improve the accuracy and robustness of the model. Through these steps, the target multi-dimensional judgment model is finally obtained, which can effectively monitor and warn the force condition of the lift and ensure the safe operation of the lift.

[0026] In the embodiment of the present application, by installing an intelligent tray at the end of the lift arm and combining it with an intelligent IoT control box, the real-time collection and transmission of the force data of the four trays is realized. A multi-dimensional judgment model including the minimum force threshold, force balance and force change rate is adopted to comprehensively evaluate the force condition of the tray, thereby improving the accuracy and reliability of abnormal detection. The force data is processed to accurately identify the motion state of the lift and the force abnormality of each tray. According to the degree of abnormality, the judgment of three levels of abnormal state, namely, slight, moderate and severe, is realized, and corresponding early warning and control strategies are adopted for different levels, effectively reducing the safety risk. By recording and analyzing the second force data, the system can dynamically optimize the early warning threshold and lifting process parameters, update the judgment model, and optimize the lifting process parameter model. By analyzing and evaluating each intelligent tray separately, the system can adapt to the needs of different models and different usage scenarios, effectively reducing the impact of data noise, improving the accuracy of abnormal detection, building a complete safety monitoring chain, and comprehensively improving the accuracy of the early warning of the force monitoring of the tray of the car lift.

[0027] In a specific embodiment, the process of executing step S102 may specifically include the following steps: (1) Setting the initial minimum force threshold of the smart pallet and calculating the force balance threshold based on the historical force data of the four smart pallets to obtain the initial force balance threshold; (2) Setting the sampling frequency of the smart pallet, and calculating the force change rate threshold based on the sampling frequency and historical force data to obtain the initial force change rate threshold; (3) Setting the conditions for judging the motion state of the lift to obtain the judgment criteria for the three states of rising, descending and stationary; (4) According to the initial minimum force threshold, the initial force balance threshold and the initial force change rate threshold, the abnormal state classification standard is formulated to obtain the judgment conditions of the three levels of abnormal states: mild, moderate and severe. The corresponding relationship between the abnormal state and the early warning response is established to obtain the early warning and control strategies for abnormal states of different levels; (5) Calculate the effective load range of the smart pallet based on the historical force data to obtain the normal working load range of each smart pallet; (6) Constructing an original judgment model based on judgment criteria, judgment conditions, normal working load range, and early warning and control strategies for different levels of abnormal conditions; (7) The weight distribution of the original judgment model is set to obtain the initial weight value of each judgment indicator in the comprehensive evaluation, and the parameters of the original judgment model are adjusted according to the initial weight value to obtain the initial multi-dimensional judgment model.

[0028] Specifically, analyze and understand the basic stress characteristics of each smart pallet. The setting of the initial minimum force threshold is mainly based on the material strength, design load and actual working environment of the smart pallet. For example, assuming that the maximum design load capacity of the smart pallet is 1,000 kg, the initial minimum force threshold is set to 10% of the design load, that is, 100 kg, to ensure that the pallet will not issue an alarm when the load is less than 100 kg, thereby avoiding unnecessary interference. The force balance threshold is calculated using the historical force data of the smart pallet. The historical data records the stress conditions of the four smart pallets under various working conditions. By performing statistical analysis on these data, the force balance threshold is calculated. The force balance threshold can be calculated using the following formula: Force balance ; in, Represents the force value of the i-th smart pallet. The formula calculates the ratio of the maximum force value to the minimum force value, thereby reflecting the degree of force imbalance. By statistically analyzing the force balance in the historical data, a reasonable threshold is determined. Assuming that this value is 1.2, it means that when the ratio of the maximum force value to the minimum force value exceeds 1.2, the system will determine that the force is unbalanced. Set the sampling frequency of the smart pallet to ensure real-time monitoring of the force changes of the pallet. The setting of the sampling frequency should take into account the working speed of the lift and the actual operation requirements. Assuming that the maximum lifting speed of the lift is 10 cm / s, in order to capture subtle force changes, the sampling frequency can be set to 10 Hz, that is, data is collected 10 times per second, thereby providing sufficient resolution to analyze the force change rate. According to the sampling frequency and historical force data, the initial force change rate threshold is calculated. The force change rate reflects the dynamic change of the force and can be obtained by first-order difference operation: ;in, Indicates the time interval The force change rate of the i-th smart tray is and are the force values ​​at the current and previous time points, respectively. A reasonable initial threshold is determined by statistically analyzing the force change rate in the historical data. For example, when the force change rate exceeds 5 kg / s, the system will issue an early warning. Subsequently, the conditions for judging the motion state of the lift are set. According to the characteristics of the force change, the motion state is divided into three states: rising, falling and static. The rising state corresponds to the situation where the force value of the pallet continues to increase, the falling state corresponds to the situation where the force value of the pallet continues to decrease, and the static state corresponds to the situation where the force value changes slightly. The specific judgment criteria can be set as follows: when the force change rate is greater than a certain positive threshold, it is judged as an rising state; when the force change rate is less than a certain negative threshold, it is judged as a falling state; when the force change rate is between the positive and negative thresholds, it is judged as a static state. Based on the above-mentioned initial minimum force threshold, initial force balance threshold and initial force change rate threshold, the abnormal state classification standard is formulated. The abnormal state is divided into three levels: mild, moderate and severe, corresponding to different risk levels. For example, when the force value exceeds the initial minimum force threshold but does not exceed 50% of the design load, it is judged as a slight abnormality; when the force value exceeds 50% of the design load but does not exceed 80% of the design load, it is judged as a moderate abnormality; when the force value exceeds 80% of the design load, it is judged as a severe abnormality. Set the corresponding early warning and control strategy according to each level of abnormal state, such as issuing a warning display for a slight abnormality, triggering an alarm for a moderate abnormality, and sending a lift stop command for a severe abnormality. According to the historical force data, the effective load range of the smart pallet is calculated. The effective load range refers to the load range that the smart pallet can safely carry under normal working conditions. By statistically analyzing the force conditions in the historical data, the normal working load range of each smart pallet is determined. For example, the normal working load range of a smart pallet is 200 to 800 kilograms. According to the set judgment criteria, judgment conditions, normal working load range and early warning and control strategies for different levels of abnormal states, the original judgment model is constructed. The original judgment model comprehensively considers various indicators. Through multi-dimensional analysis, it can comprehensively evaluate the force conditions of the pallet and make early warning and control decisions in time. The weight distribution of the original judgment model is set to determine the initial weight value of each judgment indicator in the comprehensive evaluation. The weight value setting needs to consider the importance of each indicator and the degree of influence on the overall judgment result. For example, different weights are assigned to the minimum force threshold, force balance threshold, and force change rate threshold, and the parameters of the original judgment model are adjusted according to the initial weight value to obtain the initial multi-dimensional judgment model.

[0029] In a specific embodiment, the process of executing step S103 may specifically include the following steps: (1) The intelligent IoT control box collects the first force data of the four intelligent pallets respectively to obtain four sets of original force data sets; (2) According to each set of original force data sets, the force values ​​of each smart pallet are filtered to obtain four sets of filtered force data sets; (3) Perform time series analysis on each group of filtered force data sets to obtain the force change trend of each smart pallet, and comprehensively judge the motion state of the lift based on the force change trends of the four smart pallets to obtain the current motion state of the lift; (4) Calculate the force balance of each smart pallet based on the four sets of filtered force data sets; (5) Perform differential operations on each set of filtered force data sets to obtain the force change rate of each smart pallet; (6) According to the initial minimum force threshold, initial force balance threshold and initial force change rate threshold in the initial multi-dimensional judgment model, the force value, force balance and force change rate of each smart pallet are compared respectively to obtain the abnormality score of each smart pallet corresponding to each indicator; (7) Based on the weights in the initial multi-dimensional judgment model, the abnormality scores of each smart pallet are weighted and summed to obtain the comprehensive abnormality score of each smart pallet; (8) According to the comprehensive abnormality score of each smart pallet and the abnormal state grading standard in the initial multi-dimensional judgment model, the abnormal stress state of each smart pallet is judged respectively to obtain the abnormal state level of each smart pallet; (9) Combine the lift motion status and the abnormal status level of each smart pallet to obtain a comprehensive analysis result.

[0030] Specifically, the intelligent IoT control box collects the first force data of the four intelligent pallets respectively. These data are collected in real time by sensors to obtain four sets of original force data sets. These data sets contain the force conditions of each intelligent pallet at different time points, reflecting the force changes during the entire lifting process. According to each set of original force data sets, the force value of each intelligent pallet is filtered. Filtering can use a variety of methods, such as moving average filtering or Kalman filtering, to eliminate noise and outliers in the data and obtain smoother and more reliable force data. After filtering, four sets of filtered force data sets are obtained, which can better reflect the actual force conditions of the pallets. Time series analysis is performed on each set of filtered force data sets to obtain the force change trend of each intelligent pallet. Time series analysis can reveal the laws behind the data by analyzing the trend, seasonality and periodicity of the data. By performing time series analysis on the force data of each intelligent pallet, a trend line of force change is obtained, which reflects the force changes of the pallet in different time periods. According to the force change trends of the four intelligent pallets, the motion state of the lift is comprehensively judged. The movement states of the lift include rising, falling and stationary. The current movement state of the lift is determined by analyzing the force trend of each pallet. For example, if the force values ​​of the four pallets are increasing, it can be judged that the lift is in an ascending state; if the force values ​​are decreasing, it is judged to be in a descending state; if the force values ​​do not change much, it is judged to be stationary. Based on the four sets of filtered force data sets, the force balance of each smart pallet is calculated separately. The force balance can be determined by calculating the difference between the force values ​​of each pallet. The formula is as follows: in, Represents the force value of the i-th smart pallet. This formula calculates the ratio of the maximum force value to the minimum force value, thereby reflecting the degree of force imbalance. Perform differential operations on each set of filtered force data sets to obtain the force change rate of each smart pallet. The force change rate calculation formula is: ;in, Indicates the time interval The force change rate of the i-th smart tray is and are the force values ​​at the current and previous time points, respectively. Through this calculation, the force change rate sequence of each smart pallet is obtained, reflecting the dynamic change of the force. According to the initial minimum force threshold, initial force balance threshold and initial force change rate threshold in the initial multi-dimensional judgment model, the force value, force balance and force change rate of each smart pallet are compared respectively to obtain the abnormality score of each indicator corresponding to each smart pallet. For example, the set initial minimum force threshold is 100 kg. If the force value of a pallet is less than this threshold, the abnormality score is high; the initial threshold of force balance is 1.2. If the force balance exceeds this value at a certain time point, the abnormality score increases; the initial threshold of force change rate is 5 kg / s. If the change rate exceeds this value, the corresponding score is also high. Based on the weights in the initial multi-dimensional judgment model, the abnormality scores of each smart pallet are weighted and summed to obtain the comprehensive abnormality score of each smart pallet. According to the comprehensive abnormality score of each smart pallet and the abnormal state grading standard in the initial multi-dimensional judgment model, the abnormal stress state of each smart pallet is judged separately to obtain the abnormal state level of each smart pallet. For example, if the comprehensive abnormality score of a pallet is lower than the set slight abnormality threshold, it is judged as a slight abnormality; if the score is higher than the slight abnormality threshold but lower than the moderate abnormality threshold, it is judged as a moderate abnormality; if the score exceeds the moderate abnormality threshold, it is judged as a severe abnormality. The lift motion state and the abnormal state level of each smart pallet are combined to obtain a comprehensive analysis result, which comprehensively reflects the current motion state of the lift and the abnormal stress conditions of each pallet.

[0031] In a specific embodiment, the execution step performs time series analysis on each group of filtered force data sets to obtain the force change trend of each smart tray, and comprehensively judges the motion state of the lift according to the force change trends of the four smart trays. The process of obtaining the current motion state of the lift can specifically include the following steps: (1) Perform a 5-point moving average process on each group of filtered force data sets to obtain a smoothed force data sequence, and perform a first-order difference operation on the smoothed force data sequence to obtain a force variation sequence; (2) Based on the least squares method, a linear regression is performed on the force variation series to obtain the linear equation of the force variation trend of each smart pallet; (3) According to the slope a of the linear equation of the force change trend i , set the threshold ε 1 and ε 2 , ε 1 <ε 2 , classify the force change trend of each smart pallet: when a i >ε 2 When ai <-ε 2 When |a i |≤ε 1 When , it is judged as a stationary trend; (4) Statistic the trend judgment results of the four smart pallets to obtain the frequency N of rising, falling and stationary u 、N d and N s ; (5) Set the weight coefficient W according to the frequency of trend occurrence u , W d and W s , calculate the comprehensive trend indicator CT: CT=(W u *N u +W d *N d +W s *N s ) / 3; (6) Setting the threshold CT of the comprehensive trend indicator u and CT d , compare the threshold value of the comprehensive trend indicator CT: when CT>CT u When CT <CT d When CT d ≤CT≤CT u When , the preliminary judgment result is static state; (7) Using the Kalman filter algorithm to perform state estimation on the preliminary judgment result, and obtain the corrected motion state judgment result; (8) Based on the corrected motion state judgment results, count the state sequences of the past N time points, calculate the state transition frequency, and obtain the state transition probability matrix; (9) Using the state transition probability matrix and the current state, predict the lift motion state at the next moment and obtain the current lift motion state.

[0032] Specifically, a 5-point moving average process is performed on each set of filtered force data sets. The moving average process calculates the average value of every five consecutive data points, eliminates random fluctuations and noise in the data, and obtains a smoothed force data sequence. Assume that the force data sequence is , the data sequence after moving average The calculation formula is: ; in is the smoothed Data points. Perform first-order difference operation on the smoothed force data sequence to obtain the force change sequence. The difference operation is used to calculate the change of adjacent time points. The formula is as follows: ;

[0033] in It is The force variation at each time point. Based on the least squares method, a linear regression is performed on the force variation sequence to obtain the linear equation of the force variation trend of each smart pallet. Assume that the force variation sequence is , the goal of linear regression is to find an equation of a line: ;in is the slope, indicating the trend of force change, is the time variable, is the intercept. By using the least squares method, we get the optimal and , so that the prediction error is minimized. According to the slope of the linear equation of the force change trend , set the threshold and , , classify the force change trend of each smart pallet. When , it is judged as an upward trend; when When , it is judged as a downward trend; when When , it is judged as a stationary trend; when The trend judgment results of the four smart pallets are statistically analyzed to obtain the frequency of occurrence of rising, falling, stationary and uncertain trends. , and . Set weight coefficients based on these frequencies , and , and calculate the comprehensive trend indicator , the formula is as follows: ;in, , and are the weights of rising, falling and stationary trends respectively, , and The frequency of occurrence of rising, falling and stationary trends respectively. Set the threshold of the comprehensive trend indicator and , and the comprehensive trend indicator Perform threshold comparison. When , the preliminary judgment result is rising state; when When , the preliminary judgment result is a descending state; when When , the preliminary judgment result is static state. The Kalman filter algorithm is used to estimate the state of the preliminary judgment result to obtain the corrected motion state judgment result. The Kalman filter algorithm is a recursive algorithm that can improve the accuracy and stability of state judgment through continuous updating and correction. Assume that the preliminary judgment result is , the state estimate after Kalman filtering is , updated by the following formula: ;

[0034] in, is the Kalman gain, which determines the influence of new information on state estimation. Based on the corrected motion state judgment results, the past The state sequence of each time point is calculated, and the state transition frequency is obtained to obtain the state transition probability matrix. The state transition probability matrix reflects the transition rules between different states. By analyzing these rules, future state changes can be predicted. Using the state transition probability matrix and the current state, the motion state of the lift at the next moment is predicted to obtain the current motion state of the lift. Assume that the state transition probability matrix is , the current status is , the state prediction for the next moment is , then: ; Through the above steps, the force change trend of the smart pallet can be systematically analyzed and determined, and the current movement state of the lift can be accurately predicted and determined.

[0035] In a specific embodiment, the process of respectively calculating the force balance degree of each smart pallet based on the four sets of filtered force data sets may specifically include the following steps: (1) Divide the four sets of filtered force data sets into time windows to obtain force data subsets in multiple time windows, and perform statistical analysis on the force data subsets in each time window to obtain the average force value of each smart pallet in each time window; (2) According to the average force value of each smart pallet in each time window, the force values ​​of the four smart pallets are sorted to obtain the sorting result from large to small, and the ratio of the maximum force value to the minimum force value in the sorting result is calculated to obtain the initial force balance degree of each time window; (3) Screening outliers according to the initial force balance, removing outliers beyond the preset range, obtaining a corrected force balance sequence, and performing exponential smoothing on the corrected force balance sequence to obtain a smoothed force balance sequence; (4) Based on the smoothed force balance degree sequence, the change trend of the force balance degree is analyzed to obtain the trend index of the force balance degree, and the force balance degree is predicted based on the trend index of the force balance degree to obtain the predicted force balance degree at the next moment; (5) A weighted average is taken based on the current force balance and the predicted force balance to obtain a comprehensive force balance index, and the comprehensive force balance index is normalized to obtain the force balance of each smart pallet.

[0036] Specifically, the four sets of filtered force data sets are divided into time windows, and the continuous force data are divided into several time windows of fixed length to facilitate local statistical analysis. Assume that each time window contains data points, the entire force data series Divide into subsets ,in Indicates In each time window, the average force value of each smart pallet is calculated using the following formula: ; in, Indicates The first smart pallet The average force value in the time window is Indicates Smart pallets in the time window Neidi The force value of each data point. According to the average force value of each smart pallet in each time window, the force values ​​of the four smart pallets are sorted to obtain the sorting result of the force values ​​from large to small. The ratio of the maximum force value and the minimum force value in the sorting result is calculated to obtain the initial force balance of each time window. According to the initial force balance, the outliers are screened and the outliers that exceed the preset range are eliminated. For example, the normal range of the force balance is set to [1.0, 1.5], and the outliers that exceed this range are eliminated to obtain the corrected force balance sequence. The corrected force balance sequence is subjected to exponential smoothing to eliminate random fluctuations and noise in the data. The formula for exponential smoothing is as follows: ; in, Indicates the force balance value after smoothing. Indicates the current force balance value. is the smoothing coefficient, and its value range is [0, 1]. Based on the smoothed force balance degree sequence, the change trend of the force balance degree is analyzed to obtain the trend index of the force balance degree. Trend analysis can be achieved by calculating the slope of the smoothed force balance degree. According to the trend index of the force balance degree, the force balance degree is predicted to obtain the predicted force balance degree at the next moment. The prediction can be based on linear regression or other time series prediction models. Assuming that linear regression is used, the prediction formula is as follows: ; in, is the predicted force balance, is the regression coefficient. Based on the weighted average of the current force balance and the predicted force balance, the comprehensive force balance index is obtained. The comprehensive force balance index is normalized to obtain the force balance of each smart pallet. The comprehensive force balance is converted into a value within the standard range for easy comparison and analysis.

[0037] In a specific embodiment, the execution step performs a differential operation on each group of filtered force data sets to obtain the force change rate of each smart pallet, which may specifically include the following steps: (1) Each group of filtered force data sets is sampled at equal intervals to obtain a discrete force data sequence with a fixed time interval, and a first-order difference operation is performed on the discrete force data sequence to obtain a force difference sequence at adjacent time points; (2) According to the sampling time interval and the force difference sequence, the instantaneous force change rate at each time point is calculated to obtain the original force change rate sequence, and the original force change rate sequence is median filtered to obtain the smoothed force change rate sequence; (3) Segment processing is performed on the smoothed force change rate sequence to obtain force change rate subsequences in multiple time windows, and statistical analysis is performed on the force change rate subsequences in each time window to obtain the force change rate characteristics of each time window; (4) Based on the force change rate characteristics of each time window, the linear regression method is used to fit the force change rate trend line to obtain the change trend of the force change rate, and prediction is performed based on the change trend of the force change rate to obtain the predicted force change rate sequence; (5) The smoothed force change rate sequence is weightedly fused with the predicted force change rate sequence to obtain a comprehensive force change rate sequence, and the comprehensive force change rate sequence is normalized to obtain the force change rate of each smart pallet.

[0038] Specifically, each group of filtered force data sets is sampled at equal intervals. Assume that the sampling time interval is , a discrete force data sequence with a fixed time interval is obtained by sampling at equal intervals. The continuous force data is converted into discrete data points for subsequent analysis. Assume that the original force data set is , a discrete data sequence is obtained by sampling at equal intervals ,in Indicates The time of the sampling points. Perform the first-order difference operation on the discrete force data sequence to calculate the force difference sequence of adjacent time points. The formula for the first-order difference operation is as follows: ; in, Indicates The force difference at each time point is calculated by first-order difference operation to obtain the rate of force change. The instantaneous force change rate at each time point is calculated by the force difference sequence to obtain the original force change rate sequence. The calculation formula for the instantaneous force change rate is: ; in, Indicates In order to eliminate the noise and abnormal values ​​in the force change rate sequence, the original force change rate sequence is median filtered. Median filtering is an effective denoising method. It filters the middle value in a set of data to remove abnormal fluctuations and obtain a smoothed force change rate sequence. . The smoothed force change rate sequence is segmented and divided into multiple time windows to obtain multiple force change rate subsequences. Each time window contains several data points. Through segmentation, the force changes in different time periods are analyzed. Statistical analysis is performed on the force change rate subsequences in each time window, and the statistical features of each time window are calculated. These features include mean value, standard deviation, maximum value, minimum value, etc. For example, The average value of the force change rate in the time window The formula is as follows: ;

[0039] in, Indicates The average force change rate in a time window is Indicates In the time window The force change rate of each data point. Based on the force change rate characteristics of each time window, the linear regression method is used to fit the force change rate trend line to obtain the change trend of the force change rate. The goal of linear regression is to find a straight line equation: ; in, is the slope, which indicates the trend of the rate of change of force, is the time variable, is the intercept. According to the changing trend of the force change rate, the slope of the trend line is used Make a prediction and get the predicted force change rate sequence. The prediction formula is: ; in, is the predicted force change rate. The smoothed force change rate sequence is weighted fused with the predicted force change rate sequence to obtain the comprehensive force change rate sequence. The weighted fusion formula is: ; in, represents the comprehensive force change rate, is the weighting coefficient, and its value range is [0, 1]. The comprehensive force change rate sequence is normalized and converted into values ​​within the standard range for easy comparison and analysis.

[0040] In a specific embodiment, the process of executing step S105 may specifically include the following steps: (1) Segment the second force data of each smart pallet into time windows to obtain force data subsets within multiple time windows; (2) Statistically analyzing the subset of force data within each time window to obtain the statistical characteristics of each window, and performing pattern recognition on the second force data of each smart pallet based on the statistical characteristics of each window to obtain the force pattern of each smart pallet under different working conditions; (3) Compare and analyze the stress patterns of each smart pallet under different working conditions to obtain optimized abnormality judgment criteria; (4) Based on the optimized abnormal judgment standard, the initial minimum force threshold, the initial force balance threshold, and the initial force change rate threshold in the initial multi-dimensional judgment model are dynamically adjusted to obtain the dynamically optimized early warning threshold. The dynamically optimized early warning threshold includes the target minimum force threshold, the target force balance threshold, and the target force change rate threshold; (5) Train the historical force data and the corresponding lift operation status, and combine the dynamically optimized warning threshold to obtain the lifting process parameters; (6) Reconstruct the judgment logic of the initial multi-dimensional judgment model according to the dynamically optimized warning threshold and lifting process parameters to obtain an updated judgment logic; (7) The weights of each indicator in the updated judgment logic are optimized to obtain a new weight distribution scheme. Based on the new weight distribution scheme, the parameters of the updated judgment logic are fine-tuned to obtain the target multi-dimensional judgment model.

[0041] Specifically, the second force data of each smart pallet is segmented into time windows, and the continuous force data is divided into multiple time windows of fixed length to facilitate local statistical analysis. Assume that each time window contains data points, the entire force data series Divide into subsets ,in Indicates time windows. In each time window, the statistical features of each window are calculated, including the mean, standard deviation, maximum value, minimum value, etc. According to the statistical features of each time window, the second force data of each smart pallet is subjected to pattern recognition. Pattern recognition can use methods such as cluster analysis and principal component analysis to classify the force data under different working conditions and obtain the force pattern of each smart pallet under different working conditions. For example, using the K-means clustering algorithm, the force data is divided into Classes, each representing a force mode. Compare and analyze the force modes of each smart pallet under different working conditions, and identify the typical force modes and abnormal force modes under each working condition. Through comparative analysis, determine the force characteristics under different working conditions, and optimize the abnormal judgment criteria based on this. According to the optimized abnormal judgment criteria, dynamically adjust the initial minimum force threshold, initial force balance threshold, and initial force change rate threshold in the initial multi-dimensional judgment model to obtain a dynamically optimized early warning threshold. The dynamically optimized early warning threshold includes the target minimum force threshold, target force balance threshold, and target force change rate threshold. The adjustment of these thresholds is based on the analysis of historical data and pattern recognition results, so that the judgment model is more adapted to actual working conditions. Train the historical force data and the corresponding lift operation status, and combine the dynamically optimized early warning threshold to obtain the lifting process parameters. Through machine learning algorithms, such as decision trees, support vector machines, or neural networks, train a model that can predict the force state and potential risks in the lifting process under a given dynamic early warning threshold. According to the dynamically optimized warning threshold and lifting process parameters, the judgment logic of the initial multi-dimensional judgment model is reconstructed to obtain the updated judgment logic. This includes redesigning the rules and conditions of the judgment model so that it can make accurate judgments based on the latest warning threshold and process parameters. Subsequently, the weights of each indicator in the updated judgment logic are optimized to obtain a new weight distribution scheme. Weight optimization can be achieved through weighted least squares, genetic algorithms or other optimization algorithms to ensure that the importance of each indicator in the comprehensive judgment is reasonably reflected. Based on the new weight distribution scheme, the parameters of the updated judgment logic are fine-tuned to obtain the target multi-dimensional judgment model. By fine-tuning the parameters, the accuracy and robustness of the model are improved to ensure that abnormal stress conditions can be effectively identified and warned in practical applications.

[0042] The above describes the method for monitoring and warning the force on the tray of the car lift in the embodiment of the present application. The following describes the device for monitoring and warning the force on the tray of the car lift in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of the vehicle lift tray force monitoring and early warning device includes: Installation module 201, used to install smart pallets with pressure sensing and wireless communication functions at the four ends of the lift support arm respectively, and at the same time, deploy a smart IoT control box in the lift control area to establish a wireless communication connection between the smart IoT control box and the smart pallet; Construction module 202, for the smart IoT control box to set the pallet force judgment logic of the smart pallet and to construct an initial multi-dimensional judgment model including a minimum force threshold, a force balance degree, and a force change rate; Analysis module 203, used for the intelligent IoT control box to collect the first force data transmitted by the four intelligent pallets, and analyze the force data through the initial multi-dimensional judgment model to obtain a comprehensive analysis result, which includes: the lifting machine movement state and the abnormal force state information of each intelligent pallet; Early warning module 204, used for the intelligent IoT control box to determine the early warning level and execute corresponding safety control measures according to the comprehensive analysis results. The safety control measures include issuing warning displays, triggering alarms, and sending lift control instructions; The updating module 205 is used to record and analyze the second force data of each smart pallet by the smart IoT control box after executing the corresponding safety control measures, obtain the dynamically optimized early warning threshold and lifting process parameters, and update the initial multi-dimensional judgment model through the dynamically optimized early warning threshold and lifting process parameters to obtain the target multi-dimensional judgment model.

[0043] Through the cooperation of the above components, by installing a smart tray at the end of the lift arm and combining it with the smart IoT control box, the real-time collection and transmission of the force data of the four trays are realized. A multi-dimensional judgment model including the minimum force threshold, force balance and force change rate is adopted to comprehensively evaluate the force condition of the tray, thereby improving the accuracy and reliability of abnormality detection. The force data can be processed to accurately identify the movement state of the lift and the abnormal force condition of each tray. According to the degree of abnormality, the judgment of three levels of abnormal state, namely, slight, moderate and severe, is realized, and corresponding early warning and control strategies are adopted for different levels, effectively reducing the safety risk. By recording and analyzing the second force data, the system can dynamically optimize the early warning threshold and lifting process parameters, update the judgment model, and optimize the lifting process parameter model. By analyzing and evaluating each smart tray separately, the system can adapt to the needs of different models and different usage scenarios, effectively reducing the impact of data noise, improving the accuracy of abnormality detection, building a complete safety monitoring chain, and comprehensively improving the accuracy of the early warning of the force monitoring of the tray of the car lift.

[0044] like Figure 3 As shown, Figure 3 The present application also provides a pallet force monitoring and early warning device for a car lift, the pallet force monitoring and early warning device for the car lift comprises a lift column, a lift support arm connected to the lift column; four smart pallets, including smart pallet 1, smart pallet 2, smart pallet 3 and smart pallet 4, which are respectively installed at the four ends of the lift support arm, each smart pallet has pressure sensing and wireless communication functions; an intelligent Internet of Things control box 5, which is arranged in the lift control area, and the intelligent Internet of Things control box 5 establishes a wireless communication connection with the four smart pallets to execute the steps of the pallet force monitoring and early warning method for the car lift in the above-mentioned embodiments.

[0045] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein instructions are stored in the computer, and when the instructions are executed on a computer, the computer executes the steps of the method for monitoring and warning the force on the tray of the car lift.

[0046] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0047] 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. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute 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), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0048] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for monitoring and warning the force on the tray of a car lift, characterized in that: The method comprises: Smart pallets with pressure sensing and wireless communication functions are installed at the four ends of the lift support arm respectively. At the same time, a smart IoT control box is deployed in the lift control area to establish a wireless communication connection between the smart IoT control box and the smart pallet; The smart IoT control box sets the pallet force judgment logic of the smart pallet and constructs an initial multi-dimensional judgment model including a minimum force threshold, a force balance degree, and a force change rate; The intelligent IoT control box collects the first force data transmitted by the four intelligent pallets, and analyzes the force data through the initial multi-dimensional judgment model to obtain a comprehensive analysis result, which includes: the lifting machine movement state and the abnormal force state information of each intelligent pallet; The intelligent IoT control box determines the warning level and executes corresponding safety control measures according to the comprehensive analysis results, and the safety control measures include issuing a warning display, triggering an alarm, and sending a lift control instruction; After executing the corresponding safety control measures, the intelligent IoT control box records and analyzes the second force data of each smart pallet to obtain dynamically optimized warning thresholds and lifting process parameters, and updates the initial multi-dimensional judgment model through the dynamically optimized warning thresholds and lifting process parameters to obtain the target multi-dimensional judgment model.

2. The method for monitoring and warning the force on the tray of a car lift according to claim 1, characterized in that: The smart IoT control box sets the pallet force judgment logic of the smart pallet and constructs an initial multi-dimensional judgment model including a minimum force threshold, a force balance degree, and a force change rate, including: An initial minimum force threshold of the smart pallet is set, and a force balance threshold is calculated according to historical force data of four smart pallets to obtain an initial force balance threshold; Setting a sampling frequency of the smart pallet, and calculating a force change rate threshold according to the sampling frequency and historical force data to obtain an initial force change rate threshold; The conditions for judging the movement state of the lift are set to obtain the judgment criteria for the three states of rising, falling and stationary; According to the initial minimum force threshold, the initial force balance threshold and the initial force change rate threshold, the abnormal state classification standard is formulated to obtain the judgment conditions of the three levels of abnormal states: slight, moderate and severe, and the corresponding relationship between the abnormal state and the early warning response is established to obtain the early warning and control strategies of abnormal states of different levels; Calculating the effective load range of the smart pallet according to the historical force data to obtain the normal working load range of each smart pallet; Constructing an original judgment model according to the judgment criteria, the judgment conditions, the normal working load range, and the early warning and control strategies of the abnormal states of different levels; The weight distribution of the original judgment model is set to obtain the initial weight value of each judgment indicator in the comprehensive evaluation, and the parameters of the original judgment model are adjusted according to the initial weight value to obtain the initial multi-dimensional judgment model.

3. The method for monitoring and warning the force on the tray of a car lift according to claim 2, characterized in that: The intelligent IoT control box collects the first force data transmitted by the four intelligent pallets, and analyzes the force data through the initial multi-dimensional judgment model to obtain a comprehensive analysis result, which includes: the lifting machine motion state and the abnormal force state information of each intelligent pallet, including: The smart IoT control box collects first force data of four smart pallets respectively to obtain four groups of original force data sets; According to each set of original force data sets, the force values ​​of each smart pallet are filtered to obtain four sets of filtered force data sets; Perform time series analysis on each group of filtered force data sets to obtain the force change trend of each smart pallet, and comprehensively judge the motion state of the lift based on the force change trends of the four smart pallets to obtain the current motion state of the lift; The force balance of each smart pallet is calculated based on four sets of filtered force data sets; Perform differential operations on each group of filtered force data sets to obtain the force change rate of each smart pallet; According to the initial minimum force threshold, the initial force balance threshold and the initial force change rate threshold in the initial multi-dimensional judgment model, the force value, force balance and force change rate of each smart pallet are compared respectively to obtain the abnormality score of each indicator corresponding to each smart pallet; Based on the weights in the initial multi-dimensional judgment model, the abnormality scores of each smart pallet are weighted and summed to obtain the comprehensive abnormality score of each smart pallet; According to the comprehensive abnormality score of each smart pallet and the abnormal state grading standard in the initial multi-dimensional judgment model, the abnormal stress state of each smart pallet is judged respectively to obtain the abnormal state level of each smart pallet; The lift motion status and the abnormal status level of each smart pallet are combined to obtain a comprehensive analysis result.

4. The method for monitoring and warning the force on the tray of a car lift according to claim 3 is characterized in that: The time series analysis is performed on each group of filtered force data sets to obtain the force change trend of each smart pallet, and the motion state of the lift is comprehensively judged according to the force change trends of the four smart pallets to obtain the current motion state of the lift, including: Perform 5-point moving average processing on each group of filtered force data sets to obtain a smoothed force data sequence, and perform first-order difference operation on the smoothed force data sequence to obtain a force variation sequence; Based on the least square method, linear regression is performed on the force variation series to obtain the linear equation of the force variation trend of each smart pallet; According to the slope a of the straight line equation of the force change trend i , set thresholds ε1 and ε2, ε1<ε2, and classify the force change trend of each smart pallet: when a i >ε2, it is judged as an upward trend; when a i <-ε2, it is judged as a downward trend; when |a i When |≤ε1, it is judged as a stationary trend; The trend judgment results of the four smart pallets are statistically analyzed to obtain the frequency N of rising, falling and stationary u 、N d and N s ; Set the weight coefficient W according to the frequency of trend occurrence u , W d and W s , calculate the comprehensive trend indicator CT: CT=(W u *N u +W d *N d +W s *N s ) / 3; Setting the threshold CT of the composite trend indicator u and CT d , compare the threshold value of the comprehensive trend indicator CT: when CT>CT u When CT <CT d When CT d ≤CT≤CT u When , the preliminary judgment result is static state; The Kalman filter algorithm is used to perform state estimation on the preliminary judgment result to obtain a corrected motion state judgment result; Based on the corrected motion state judgment result, count the state sequence of the past N time points, calculate the state transition frequency, and obtain the state transition probability matrix; The state transition probability matrix and the current state are used to predict the motion state of the lift at the next moment to obtain the current motion state of the lift.

5. The method for monitoring and warning the force on the tray of a car lift according to claim 4, characterized in that: The force balance degree of each smart pallet is calculated based on the four sets of filtered force data sets, including: The four groups of filtered force data sets are divided into time windows to obtain force data subsets in multiple time windows, and statistical analysis is performed on the force data subsets in each time window to obtain the average force value of each smart pallet in each time window; According to the average force value of each smart pallet in each time window, the force values ​​of the four smart pallets are sorted to obtain the sorting result of the force values ​​from large to small, and the ratio of the maximum force value and the minimum force value in the sorting result is calculated to obtain the initial force balance degree of each time window; The abnormal values ​​are screened according to the initial force balance degree, and the abnormal values ​​beyond the preset range are eliminated to obtain a corrected force balance degree sequence, and the corrected force balance degree sequence is subjected to exponential smoothing to obtain a smoothed force balance degree sequence; Based on the smoothed force balance degree sequence, the change trend of the force balance degree is analyzed to obtain the trend index of the force balance degree, and the force balance degree is predicted according to the trend index of the force balance degree to obtain the predicted force balance degree at the next moment; Based on the weighted average of the current force balance and the predicted force balance, a comprehensive force balance index is obtained, and the comprehensive force balance index is normalized to obtain the force balance of each smart pallet.

6. The method for monitoring and warning the force on the tray of a car lift according to claim 5, characterized in that: The differential operation is performed on each group of filtered force data sets to obtain the force change rate of each smart pallet, including: Each group of filtered force data sets is sampled at equal intervals to obtain a discrete force data sequence with a fixed time interval, and a first-order difference operation is performed on the discrete force data sequence to obtain a force difference sequence at adjacent time points; According to the sampling time interval and the force difference sequence, the instantaneous force change rate at each time point is calculated to obtain the original force change rate sequence, and the original force change rate sequence is subjected to median filtering to obtain a smoothed force change rate sequence; The smoothed force change rate sequence is segmented to obtain force change rate subsequences in multiple time windows, and statistical analysis is performed on the force change rate subsequences in each time window to obtain the force change rate characteristics of each time window; Based on the force change rate characteristics of each time window, the force change rate trend line is fitted using the linear regression method to obtain the change trend of the force change rate, and prediction is made based on the change trend of the force change rate to obtain the predicted force change rate sequence; The smoothed force change rate sequence and the predicted force change rate sequence are weightedly fused to obtain a comprehensive force change rate sequence, which is then normalized to obtain the force change rate of each smart pallet.

7. The method for monitoring and warning the force on the tray of a car lift according to claim 6, characterized in that: After executing the corresponding safety control measures, the smart IoT control box records and analyzes the second force data of each smart pallet to obtain a dynamically optimized warning threshold and lifting process parameters, and updates the initial multi-dimensional judgment model through the dynamically optimized warning threshold and lifting process parameters to obtain a target multi-dimensional judgment model, including: Performing time window segmentation on the second force data of each smart pallet to obtain force data subsets within multiple time windows; Perform statistical analysis on the force data subsets in each time window to obtain the statistical characteristics of each window, and perform pattern recognition on the second force data of each smart pallet according to the statistical characteristics of each window to obtain the force mode of each smart pallet under different working conditions; Compare and analyze the stress patterns of each smart pallet under different working conditions to obtain optimized abnormality judgment criteria; Based on the optimized abnormal judgment standard, the initial minimum force threshold, the initial force balance threshold and the initial force change rate threshold in the initial multi-dimensional judgment model are dynamically adjusted to obtain a dynamically optimized early warning threshold, wherein the dynamically optimized early warning threshold includes a target minimum force threshold, a target force balance threshold and a target force change rate threshold; The historical force data and the corresponding lift operation status are trained, and the lifting process parameters are obtained by combining the dynamically optimized warning thresholds; According to the dynamically optimized warning threshold and lifting process parameters, the judgment logic of the initial multi-dimensional judgment model is reconstructed to obtain an updated judgment logic; The weights of each indicator in the updated judgment logic are optimized to obtain a new weight distribution scheme, and based on the new weight distribution scheme, the parameters of the updated judgment logic are fine-tuned to obtain the target multi-dimensional judgment model.

8. A vehicle lift tray force monitoring and early warning device, characterized in that: The device is used to implement the method for monitoring and warning the force on the tray of the car lift according to any one of claims 1 to 7, and comprises: An installation module is used to install smart pallets with pressure sensing and wireless communication functions at the four ends of the lift support arm, and at the same time, a smart IoT control box is deployed in the lift control area to establish a wireless communication connection between the smart IoT control box and the smart pallet; A construction module, used for the smart IoT control box to set the pallet force judgment logic of the smart pallet and to construct an initial multi-dimensional judgment model including a minimum force threshold, a force balance degree and a force change rate; An analysis module, used for the smart IoT control box to collect the first force data transmitted by the four smart pallets, and to analyze the force data through the initial multi-dimensional judgment model to obtain a comprehensive analysis result, wherein the comprehensive analysis result includes: the lifting machine motion state and the abnormal force state information of each smart pallet; An early warning module, used for the intelligent IoT control box to determine the early warning level and execute corresponding safety control measures according to the comprehensive analysis results, and the safety control measures include issuing a warning display, triggering an alarm and sending a lift control instruction; An updating module is used for recording and analyzing the second force data of each smart pallet by the smart IoT control box after executing corresponding safety control measures, obtaining dynamically optimized early warning thresholds and lifting process parameters, and updating the initial multi-dimensional judgment model through the dynamically optimized early warning thresholds and lifting process parameters to obtain a target multi-dimensional judgment model.

9. A vehicle lift tray force monitoring and early warning device, characterized in that: A method for monitoring and warning the force on the pallet of a car lift as described in any one of claims 1 to 7, wherein the device comprises: a lift column, a lift support arm connected to the lift column; four smart pallets, respectively installed at the four ends of the lift support arm, each smart pallet having pressure sensing and wireless communication functions; and a smart Internet of Things control box, which is arranged in the lift control area, and the smart Internet of Things control box establishes a wireless communication connection with the four smart pallets.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the method for monitoring and warning the force on the tray of the car lift according to any one of claims 1 to 7 is implemented.

Images