An elevator fault detection and alarm system and method

Abstract

The present invention discloses an elevator fault detection and alarm system and method. The present invention relates to the technical field of elevator fault detection, and includes an elevator data acquisition module, an elevator load and buffer state detection module, and an elevator load and buffer state analysis module; the elevator data acquisition module uses a load sensor and a linear displacement sensor to monitor the deformation of the leaf spring and the displacement of the internal measuring element of the linear displacement sensor, and accurately collects elevator load and buffer compression data; the elevator load and buffer state detection module analyzes the collected data, learns the normal operation mode of the elevator, evaluates the buffer state, and constructs a detection model based on the genetic wavelet neural network algorithm; the elevator load and buffer state analysis module quickly determines whether the elevator is overloaded and whether the buffer is working properly according to the model output, ensures the safe operation of the elevator, reduces the risk of faults, and provides real-time and intelligent fault warnings.

Description

Technical Field

[0001] The present invention relates to the technical field of elevator fault detection, and particularly to an elevator fault detection and alarm system and method. Background Technique

[0002] With the acceleration of the urbanization process and the continuous increase of high-rise buildings, elevators, as the main means of vertical transportation, their safety and reliability have attracted more and more attention. Traditional elevator fault detection and alarm methods often rely on manual inspections and passive responses after faults. This method is not only inefficient but also difficult to prevent accidents. Therefore, the application of elevator fault detection and alarm systems has become particularly important. At present, significant progress has been made in elevator fault detection and alarm systems. By applying advanced technologies such as the Internet of Things, cloud computing, and artificial intelligence, the system can achieve real-time monitoring of elevator status, data collection, and fault early warning. Once an elevator fails, the system can immediately notify the maintenance personnel and provide accurate fault information, thus shortening the fault response time and improving the maintenance efficiency. In addition, with the development of intelligent automation technology, elevator fault detection and alarm systems can also achieve remote troubleshooting and preventive maintenance. Through pre-installed advanced instrument equipment, the system completes the remote and rapid detection and troubleshooting, timely discovers potential fault hazards in the elevator, and processes them in advance.

[0003] Although the current elevator fault detection and alarm systems have been widely applied, they still face safety hazards caused by delayed overload display and damaged buffers. Delayed overload display may cause the elevator to operate overloaded, increasing the risk of accidents; while damaged buffers may not provide effective buffering when the elevator falls, endangering the lives of passengers. To solve the above problems, by collecting and analyzing elevator load data and buffer status data, an elevator load and buffer status detection model based on the genetic wavelet neural network algorithm is established to deeply learn the normal operation mode of the elevator to accurately judge the operating state of the elevator. This model can process a large amount of data in real time, quickly identify the overload situation of the elevator and the health status of the buffer, and thus issue early warnings in a timely manner to effectively avoid potential safety risks. This innovative method not only improves the accuracy and efficiency of elevator fault detection but also provides a strong guarantee for the safe operation of the elevator, allowing passengers to feel more at ease while enjoying the convenience. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides an elevator fault detection and alarm system, which solves the problems in the above background technique.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: An elevator fault detection and alarm system includes an elevator data collection module, an elevator load and buffer status detection module, and an elevator load and buffer status analysis module;

[0006] The elevator data acquisition module includes a load sensor and a linear displacement sensor, and collects elevator load data and buffer compression amount data by detecting the displacement generated by the deformation of the leaf spring and the internal measuring element of the linear displacement sensor.

[0007] The elevator load and buffer state detection module analyzes the elevator load data and buffer compression amount data, learns the normal operation mode of the elevator, judges the buffer state, and establishes an elevator load and buffer state detection model based on the genetic wavelet neural network algorithm.

[0008] The elevator load and buffer state analysis module judges whether the elevator is overloaded and whether the buffer is in a normal working state based on the output results of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm.

[0009] Furthermore, the process of the elevator data acquisition module collecting elevator load data through the load sensor includes:

[0010] The load sensor is installed at the bottom of the elevator car. When the elevator bears weight, the load sensor detects the deformation of the leaf spring caused by the weight change. The deformation then generates a displacement amount. The load sensor measures this displacement amount and converts it into an output of voltage change through an internal circuit. This voltage change is proportional to the elevator load. Therefore, the elevator load is calculated by measuring the voltage change. The elevator data acquisition module is responsible for receiving the electrical signal converted by the load sensor and further processing it into a digital signal.

[0011] Furthermore, the process of the elevator data acquisition module monitoring the compression amount of the buffer through the linear displacement sensor includes:

[0012] The linear displacement sensor monitors the buffer compression amount data based on the principle of precise perception and conversion of displacement by its internal measuring element. The internal measuring element of it undergoes corresponding displacement changes as the buffer is compressed. This displacement change is captured by the circuit inside the sensor and converted into an electrical signal. This electrical signal is proportional to the buffer compression amount. The elevator data acquisition module is responsible for receiving the electrical signal converted by the linear displacement sensor and further processing it into a digital signal.

[0013] Furthermore, the process of the elevator load and buffer state detection module training the input layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by analyzing the elevator load data and buffer compression amount data includes:

[0014] Clean and organize the collected elevator load data and buffer compression data, remove invalid data and outliers, normalize the data, and establish a genetic wavelet neural network model. The genetic wavelet neural network model includes an input layer, a hidden layer, and an output layer. The input layer is responsible for receiving the elevator load data and buffer compression data and converting the continuous load data and buffer compression data into vector representations.

[0015] Furthermore, the process of training the hidden layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by the elevator load and buffer state detection module through non-linear transformation and feature extraction of the input data includes:

[0016] The hidden layer of the genetic wavelet neural network model uses the Morlet wavelet function as the activation function, takes the wavelet coefficients obtained through wavelet analysis as the input of the hidden layer, captures the high-frequency and low-frequency components in the collected data, performs multi-scale analysis on the input data, and thus extracts the features that are discriminative for the elevator operating state and buffer state. Then, the hidden layer uses the genetic algorithm to optimize the weights and thresholds of the model, randomly generates a set of combinations of neural network weights and thresholds as the initial population, configures the weights and thresholds of each individual to the neural network, and inputs the elevator load data and buffer compression data for training. According to the error between the fault detection result and the actual fault situation, calculate the fitness of each individual. According to the fitness value, select the individuals with high fitness for replication, retain their excellent features in the next generation, perform selection and replication operations through the tournament selection method, partially exchange the weights and thresholds of two individuals to generate new offspring, randomly change the values of some weights and thresholds to increase the diversity of the population, repeat the above genetic algorithm steps, continuously iterate and optimize the weights and thresholds of the neural network. When the iteration process ends, output the individual with the highest fitness as the optimal solution to obtain the optimal configuration of neural network weights and thresholds. In addition, through the interaction of each neuron in the hidden layer, fuse the extracted feature information to form a higher-level feature representation and output it to the output layer for the final judgment.

[0017] Furthermore, the process of training the output layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by the elevator load and buffer state detection module to learn the normal operation mode of the elevator and judge the buffer state includes:

[0018] The output layer of the genetic wavelet neural network model receives the output data from the hidden layer and performs fault judgment according to the preset judgment rules. If the elevator load exceeds the preset elevator load threshold, it is judged that the elevator is overloaded. If the elevator load does not exceed the preset elevator load threshold, it is judged that the elevator load is normal. If the buffer compression amount exceeds the compression threshold, it is judged that the buffer is damaged. If the buffer compression amount is lower than the compression threshold, it is judged that the buffer is in a normal state.

[0019] Further, the process of the elevator load and buffer state analysis module judging whether the elevator is overloaded and whether the buffer is in a normal working state based on the output results of the elevator load and buffer state detection model using the genetic wavelet neural network algorithm includes:

[0020] For large passenger elevators, the maximum load is set at 1600 kg. When the detected actual load exceeds 1600 kg, the system determines that it is overloaded, gives an overload prompt on the elevator display screen, prevents the elevator from closing the door and starting, and ensures the safety of passengers in case of overload. When the detected actual load is lower than 1600 kg, the system determines that the load is normal and the elevator operates normally;

[0021] The compression stroke of the elevator buffer is set between 200 and 250 mm. When the system detects that the buffer compression amount exceeds 250 mm, the system determines that the buffer state is abnormal and sends the buffer fault information to the elevator maintenance personnel for timely repair and handling. When the system detects that the buffer compression amount is within 250 mm, the system determines that the buffer state is normal and the elevator operates normally.

[0022] Further, an elevator fault detection and alarm method includes the following steps:

[0023] S1. Respectively detect the displacement amounts generated by the deformation of the leaf spring and the internal measuring element of the linear displacement sensor through the load sensor and the linear displacement sensor, and collect the elevator load data and the buffer compression amount data;

[0024] S2. Establish an elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by analyzing the elevator load data and the buffer compression amount data;

[0025] S3. Based on the output results of the elevator load and buffer state detection model using the genetic wavelet neural network algorithm, judge whether the elevator is overloaded and whether the buffer is in a normal working state.

[0026] The present invention provides an elevator fault detection and alarm system and method. It has the following beneficial effects:

[0027] First, through the accurate data acquisition of load sensors and linear displacement sensors, the system can obtain accurate information on elevator load and buffer compression in real time, providing a reliable data basis for subsequent fault detection. Secondly, the elevator load and buffer state detection module uses the genetic wavelet neural network algorithm to deeply analyze the collected data, learn the normal operation mode of the elevator, and establish a detection model based on this. This model not only improves the accuracy of fault detection but also can adapt to various changes during elevator operation, enhancing the adaptability and robustness of the system. Finally, according to the output results of the detection model, the elevator load and buffer state analysis module can quickly determine whether the elevator is overloaded and whether the buffer is in a normal working state. Once an abnormal situation is detected, the system can immediately issue an alarm signal and take timely measures to avoid potential safety risks, effectively ensuring the safety of passengers' lives and the normal operation of the elevator. In summary, through accurate data acquisition, intelligent algorithm analysis, and timely fault alarm, this system significantly improves the accuracy and efficiency of elevator fault detection, providing strong technical support for the safe operation of elevators. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a block diagram of an elevator fault detection and alarm system according to an embodiment of the present application.

[0029] Figure 2 It is a flowchart of an elevator fault detection and alarm method according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0030] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present invention.

[0031] As Figure 1 and Figure 2 shown, the present invention provides a technical solution: an elevator fault detection and alarm system, including an elevator data acquisition module, an elevator load and buffer state detection module, and an elevator load and buffer state analysis module;

[0032] The elevator data acquisition module includes a load sensor and a linear displacement sensor, and collects elevator load data and buffer compression data by detecting the displacement generated by the deformation of the leaf spring and the internal measuring element of the linear displacement sensor;

[0033] The elevator load and buffer state detection module analyzes the elevator load data and buffer compression data, learns the normal operation mode of the elevator, judges the buffer state, and establishes an elevator load and buffer state detection model based on the genetic wavelet neural network algorithm;

[0034] The elevator load and buffer state analysis module judges whether the elevator is overloaded and whether the buffer is in a normal working state based on the output results of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm.

[0035] The process of the elevator data acquisition module collecting elevator load data through the load sensor includes:

[0036] The load sensor is installed at the bottom of the elevator car. When the elevator bears weight, the load sensor detects the deformation of the leaf spring caused by the weight change. The deformation then generates a displacement. The load sensor measures this displacement and uses the internal circuit to convert it into an output of voltage change. This voltage change is proportional to the elevator load. Therefore, the elevator load is calculated by measuring the voltage change. The elevator data acquisition module is responsible for receiving the electrical signal converted by the load sensor and further processing it into a digital signal.

[0037] The process of the elevator data acquisition module monitoring the buffer compression through the linear displacement sensor includes:

[0038] The linear displacement sensor monitors the buffer compression data based on the principle of precise perception and conversion of displacement by its internal measuring element. The internal measuring element of it undergoes corresponding displacement changes as the buffer compresses. This displacement change is captured by the circuit inside the sensor and converted into an electrical signal. This electrical signal is proportional to the buffer compression. The elevator data acquisition module is responsible for receiving the electrical signal converted by the linear displacement sensor and further processing it into a digital signal.

[0039] The process of the elevator load and buffer state detection module training the input layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by analyzing the elevator load data and buffer compression data includes:

[0040] Clean and organize the collected elevator load data and buffer compression data, remove invalid data and outliers, normalize the data, and establish a genetic wavelet neural network model. The genetic wavelet neural network model includes an input layer, a hidden layer, and an output layer. The input layer is responsible for receiving the elevator load data and buffer compression data and converting the continuous load data and buffer compression data into vector representations.

[0041] The process of training the hidden layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by the elevator load and buffer state detection module through non - linear transformation and feature extraction of the input data includes:

[0042] The hidden layer of the genetic wavelet neural network model uses the Morlet wavelet function as the activation function. Taking the wavelet coefficients obtained through wavelet analysis as the input of the hidden layer, it captures the high - frequency and low - frequency components in the collected data, conducts multi - scale analysis on the input data, and thus extracts features that are discriminative for the elevator operation state and buffer state. Then, the hidden layer uses the genetic algorithm to optimize the weights and thresholds of the model. Randomly generate a set of combinations of neural network weights and thresholds as the initial population, configure the weights and thresholds of each individual to the neural network, and input the elevator load data and buffer compression data for training. Calculate the fitness of each individual according to the error between the fault detection result and the actual fault situation. Select the individuals with high fitness for replication according to the fitness value, retain their excellent features in the next generation, and perform selection and replication operations through the tournament selection method. Partially exchange the weights and thresholds of two individuals to generate new offspring, randomly change the values of some weights and thresholds to increase the diversity of the population. Repeat the above genetic algorithm steps, continuously iterate and optimize the weights and thresholds of the neural network. When the iteration process ends, output the individual with the highest fitness as the optimal solution to obtain the optimal configuration of neural network weights and thresholds. In addition, through the interaction of each neuron in the hidden layer, the extracted feature information is fused to form a higher - level feature representation and output to the output layer for final judgment.

[0043] The process of the elevator load and buffer state detection module learning the normal operation mode of the elevator, judging the buffer state, and training the output layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm includes:

[0044] The output layer of the genetic wavelet neural network model receives the output data of the hidden layer and makes a fault judgment according to the preset judgment rules. If the elevator load exceeds the preset elevator load threshold, it is judged that the elevator is overloaded. If the elevator load does not exceed the preset elevator load threshold, it is judged that the elevator load is normal. If the buffer compression exceeds the compression threshold, it is judged that the buffer is damaged. If the buffer compression is lower than the compression threshold, it is judged that the buffer state is normal.

[0045] The process of the elevator load and buffer state analysis module judging whether the elevator is overloaded and whether the buffer is in a normal working state based on the output result of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm includes:

[0046] For large passenger elevators, the maximum load is set at 1600 kg. When the detected actual load exceeds 1600 kg, the system determines that it is overloaded. An overload prompt is given on the elevator display screen to prevent the elevator from closing and starting, ensuring the safety of passengers in case of overload. When the detected actual load is lower than 1600 kg, the system determines that the load is normal and the elevator operates normally;

[0047] The compression stroke of the elevator buffer is set between 200 and 250 millimeters. When the system detects that the compression amount of the buffer exceeds 250 millimeters, the system determines that the buffer status is abnormal and sends the buffer fault information to the elevator maintenance personnel for timely repair and handling. When the system detects that the compression amount of the buffer is within 250 millimeters, the system determines that the buffer status is normal and the elevator operates normally.

[0048] It should be noted that since the load data changes with factors such as the number of passengers and the weight of goods, and the buffer compression amount data is also affected by various factors such as elevator speed and braking impact, the wavelet analysis part of the genetic wavelet neural network algorithm is used to perform multi-scale analysis on the data. The data is decomposed into different frequency components through wavelet functions, effectively capturing the high-frequency components (the sharp change part of the load data when the elevator is suddenly overloaded) and low-frequency components (the slow fluctuation part of the normal load of the elevator) in the data, so as to extract the discriminative features for the elevator operation status and buffer status. The genetic algorithm part of the genetic wavelet neural network algorithm can globally optimize the weights and thresholds of the model. By simulating the biological evolution process, an initial population (a combination of a set of neural network weights and thresholds) is randomly generated. Through selection, crossover, and mutation operations, continuous iterative optimization is carried out. After multiple generations of evolution, a combination of weights and thresholds with higher fitness is found. During the training process of the genetic wavelet neural network algorithm, it learns the normal operation mode of the elevator, including the reasonable change range of the load and the change law of the compression amount when the buffer works normally. Furthermore, in actual operation, the current data is compared with the normal mode in real time. If the elevator operation status deviates from the normal mode (the compression amount of the buffer is abnormal due to long-term use and wear), it can be quickly identified and an alarm is given;

[0049] In addition, compared with the traditional method based solely on threshold judgment, the genetic wavelet neural network algorithm can more accurately judge whether the elevator is overloaded and whether the buffer is working properly. In the buffer status detection, it can effectively distinguish between normal wear and faults of the buffer. Moreover, in the actual elevator operation environment, the multi-scale analysis and feature extraction ability of the genetic wavelet neural network algorithm for data filters out the noise interference caused by electromagnetic interference resulting in instantaneous fluctuations in sensor data, maintaining the stable operation of the model.

[0050] An elevator fault detection and alarm method includes the following steps:

[0051] S1. Detect the displacement generated by the deformation of the leaf spring and the internal measuring element of the linear displacement sensor through a load sensor and a linear displacement sensor respectively, and collect the elevator load data and the buffer compression data;

[0052] S2. Establish an elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by analyzing the elevator load data and the buffer compression data;

[0053] S3. Judge whether the elevator is overloaded and whether the buffer is in a normal working state based on the output results of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm.

[0054] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation. An element defined by the statement "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or device comprising said element.

[0055] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An elevator fault detection and alarm system, characterized in that: It includes an elevator data acquisition module, an elevator load and buffer state detection module, and an elevator load and buffer state analysis module; The elevator data acquisition module includes a load sensor and a linear displacement sensor. By detecting the displacement generated by the deformation of the leaf spring and the internal measuring element of the linear displacement sensor, it acquires elevator load data and buffer compression data; The elevator load and buffer state detection module analyzes the elevator load data and buffer compression data, learns the normal operation mode of the elevator, judges the buffer state, and establishes an elevator load and buffer state detection model based on the genetic wavelet neural network algorithm; The process of the elevator load and buffer state detection module training the input layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by analyzing the elevator load data and buffer compression data includes: Clean and sort the collected elevator load data and buffer compression data, remove invalid data and outliers, normalize the data, and establish a genetic wavelet neural network model. The genetic wavelet neural network model includes an input layer, a hidden layer, and an output layer. The input layer is responsible for receiving the elevator load data and buffer compression data, and converting the continuous load data and buffer compression data into vector representations; The process of the elevator load and buffer state detection module training the hidden layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by performing non-linear transformation and feature extraction on the input data includes: The hidden layer of the genetic wavelet neural network model uses the Morlet wavelet function as the activation function. The wavelet coefficients obtained through wavelet analysis are used as the input of the hidden layer to capture the high-frequency and low-frequency components in the collected data, perform multi-scale analysis on the input data, and thus extract features that are discriminative for the elevator operation state and buffer state. Then, the hidden layer uses the genetic algorithm to optimize the weights and thresholds of the model. Randomly generate a set of combinations of neural network weights and thresholds as the initial population, configure the weights and thresholds of each individual to the neural network, and input the elevator load data and buffer compression data for training. Calculate the fitness of each individual according to the error between the fault detection result and the actual fault situation. According to the fitness value, select individuals with high fitness for replication, retain their excellent features in the next generation, perform selection and replication operations through the tournament selection method, partially exchange the weights and thresholds of two individuals to generate new offspring, randomly change the values of some weights and thresholds to increase the diversity of the population, repeat the above genetic algorithm steps, continuously iterate and optimize the weights and thresholds of the neural network. When the iteration process ends, output the individual with the highest fitness as the optimal solution to obtain the optimal neural network weight and threshold configuration. In addition, through the interaction of each neuron in the hidden layer, the extracted feature information is fused to form a higher-level feature representation and output to the output layer for final judgment; The elevator load and buffer state detection module learns the normal operation mode of the elevator, judges the buffer state, and the process of training the output layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm includes: The output layer of the genetic wavelet neural network model receives the output data of the hidden layer and performs fault judgment according to the preset judgment rules. If the elevator load exceeds the preset elevator load threshold, it is judged that the elevator is overloaded. If the elevator load does not exceed the preset elevator load threshold, it is judged that the elevator load is normal. If the buffer compression amount exceeds the compression threshold, it is judged that the buffer is damaged. If the buffer compression amount is lower than the compression threshold, it is judged that the buffer state is normal; The elevator load and buffer state analysis module judges whether the elevator is overloaded and whether the buffer is in a normal working state based on the output results of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm.

2. The elevator fault detection and alarm system according to claim 1, characterized in that: The process of the elevator data acquisition module collecting elevator load data through the load sensor includes: The load sensor is installed at the bottom of the elevator car. When the elevator bears weight, the load sensor detects the deformation of the leaf spring caused by the weight change. The deformation then generates a displacement amount. The load sensor measures this displacement amount and converts it into an output of voltage change through an internal circuit. This voltage change is proportional to the elevator load. Therefore, the elevator load is calculated by measuring the voltage change. The elevator data acquisition module is responsible for receiving the electrical signal converted by the load sensor and further processing it into a digital signal.

3. The elevator fault detection and alarm system according to claim 2, characterized in that: The process of the elevator data acquisition module monitoring the compression amount of the buffer through the linear displacement sensor includes: The linear displacement sensor monitors the buffer compression amount data based on the principle of precise perception and conversion of displacement by its internal measuring element. The internal measuring element of it undergoes corresponding displacement changes as the buffer is compressed. This displacement change is captured by the circuit inside the sensor and converted into an electrical signal. This electrical signal is proportional to the buffer compression amount. The elevator data acquisition module is responsible for receiving the electrical signal converted by the linear displacement sensor and further processing it into a digital signal.

4. The elevator fault detection and alarm system according to claim 3, characterized in that: The process of the elevator load and buffer state analysis module judging whether the elevator is overloaded and whether the buffer is in a normal working state based on the output results of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm includes: For large passenger elevators, the maximum load is set to 1600 kg. When it is detected that the actual load exceeds 1600 kg, the system determines that it is overloaded, and an overload prompt is given on the elevator display screen to prevent the elevator from closing the door and starting, ensuring the safety of passengers in case of overload. When it is detected that the actual load is lower than 1600 kg, the system determines that the load is normal and the elevator operates normally; Set the compression stroke of the elevator buffer between 200 and 250 millimeters. When the system detects that the compression amount of the buffer exceeds 250 millimeters, the system determines that the buffer is in an abnormal state and sends the buffer fault information to the elevator maintenance personnel for timely repair and handling. When the system detects that the compression amount of the buffer is within 250 millimeters, the system determines that the buffer is in a normal state and the elevator operates normally.

5. An elevator fault detection and alarm method, characterized in that: It includes the following steps: S1. Respectively detect the displacement amounts generated by the deformation of the leaf spring and the internal measuring element of the linear displacement sensor through the load sensor and the linear displacement sensor, and collect the elevator load data and the buffer compression amount data. S2. By analyzing the elevator load data and the buffer compression amount data, establish an elevator load and buffer state detection model based on the genetic wavelet neural network algorithm. The process of training the input layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by the elevator load and buffer state detection module through analyzing the elevator load data and the buffer compression amount data includes: Clean and sort the collected elevator load data and buffer compression amount data, remove invalid data and outliers, normalize the data, and establish a genetic wavelet neural network model. The genetic wavelet neural network model includes an input layer, a hidden layer, and an output layer. The input layer is responsible for receiving the elevator load data and the buffer compression amount data, and converting the continuous load data and buffer compression amount data into vector representations. The process of training the hidden layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm by the elevator load and buffer state detection module through performing nonlinear transformation and feature extraction on the input data includes: The hidden layer of the genetic wavelet neural network model uses the Morlet wavelet function as the activation function, takes the wavelet coefficients obtained through wavelet analysis as the input of the hidden layer, captures the high-frequency and low-frequency components in the collected data, conducts multi-scale analysis on the input data, and thus extracts the discriminative features for the elevator operation state and buffer state. Then, the hidden layer uses the genetic algorithm to optimize the weights and thresholds of the model, randomly generates a set of combinations of neural network weights and thresholds as the initial population, configures the weights and thresholds of each individual to the neural network, and inputs the elevator load data and buffer compression data for training. According to the error between the fault detection result and the actual fault situation, calculates the fitness of each individual. According to the fitness value, selects the individuals with high fitness for replication, retains their excellent features in the next generation, and performs selection and replication operations through the tournament selection method. Partially exchanges the weights and thresholds of two individuals to generate new offspring, randomly changes the values of some weights and thresholds to increase the diversity of the population, repeats the above genetic algorithm steps, continuously iteratively optimizes the weights and thresholds of the neural network. When the iteration process ends, outputs the individual with the highest fitness as the optimal solution to obtain the optimal configuration of neural network weights and thresholds. In addition, through the interaction of each neuron in the hidden layer, fuses the extracted feature information to form a higher-level feature representation and outputs it to the output layer for final judgment; The elevator load and buffer state detection module learns the normal operation mode of the elevator and judges the buffer state. The process of training the output layer of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm includes: The output layer of the genetic wavelet neural network model receives the output data of the hidden layer and conducts fault judgment according to the preset judgment rules. If the elevator load exceeds the preset elevator load threshold, it is judged that the elevator is overloaded. If the elevator load does not exceed the preset elevator load threshold, it is judged that the elevator load is normal. If the buffer compression exceeds the compression threshold, it is judged that the buffer is damaged. If the buffer compression is lower than the compression threshold, it is judged that the buffer state is normal; S3. Based on the output result of the elevator load and buffer state detection model based on the genetic wavelet neural network algorithm, judge whether the elevator is overloaded and whether the buffer is in a normal working state.

Images