Embedded integrated control system of tower crane hoisting mechanism

Abstract

The application relates to a tower crane hoisting mechanism embedded integrated control system. The application designs a stress sensor, an ultrasonic detector and a vibration detector to carry out real-time monitoring on a hoisting process. The stress sensor can detect the actual stress condition of a hoisting arm in a running process, and the vibration waveform detected by the vibration detector reflects the load condition on one hand and whether various resonances exist in the running process on the other hand. The ultrasonic sensor can obtain the tension degree of a steel cable in the running process, and can comprehensively detect the hoisting process. After the load coefficient is obtained, the gravity center changes constantly in the hoisting process, and the stress of the hoisting arm changes constantly. The maximum deformation amount of the hoisting arm is taken as a parameter to calculate the actual maximum running speed, so that safety, stability and efficiency are considered to a larger extent.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of crane engineering, in particular to a tower crane hoisting mechanism embedded integrated control system. BACKGROUND

[0002] The hoisting mechanism, also known as the lifting mechanism, is an essential basic component of the crane for lifting and lowering operation. The hoisting mechanism is driven by the hoisting motor through the coupling to rotate the drum through the hollow shaft of the reducer, so as to drive the hook device on the steel wire rope / cable around the drum to rise or fall. Different hoisting heights may require corresponding couplings and support frames. The performance of the hoisting mechanism will directly affect the working performance of the whole crane.

[0003] Application No. CN201410606335.8 discloses a hoisting mechanism control method, control device and hoisting mechanism. The hoisting mechanism control method includes: receiving force information reflecting the force condition of the traction rope; determining whether the traction rope is in a slack state according to the force condition of the traction rope; if the traction rope is in a slack state, control the traction rope to be tensioned; if the traction rope is in a tensioned state, control the hoisting mechanism to continue working.

[0004] Application No. CN202110163496.4 discloses a tower crane intelligent control system, belonging to the field of building construction engineering machinery. The control system includes a control center, a wireless communication module, a cloud server, a mobile control terminal, a high-definition video monitoring module, a hook intelligent obstacle avoidance module, a data acquisition and transmission module, an action execution module and an alarm module. The cloud server, mobile control terminal, high-definition video monitoring module and hook intelligent obstacle avoidance module are connected to the control center through the wireless communication module, and the data acquisition and transmission module, action execution module and alarm module are connected to the control center.

[0005] Setting the monitoring module as described above can achieve certain monitoring effect, but the rope detection detection effect is single, and the video detection method has low detection accuracy on the one hand and needs to set a large number of monitoring during the detection process, resulting in a large amount of detection data and the need for high system configuration. SUMMARY

[0006] In view of the above, to solve the above problems, a tower crane hoisting mechanism embedded integrated control system is provided, which includes a hoisting controller, a computing module, a hoisting motor, a stress sensor, a vibration detector, an ultrasonic detector and an upper computer.

[0007] The hoisting controller embedded integrated architecture design, the calculation module, the hoisting motor, the stress sensor, the ultrasonic detector and the vibration detector are connected to the hoisting controller, the hoisting motor is used to drive the hoisting module to rise and fall, so as to realize the hoisting work of the crane; the stress sensor is arranged on the hoisting arm and is used to detect the deformation condition of the hoisting arm when carrying the goods and send the deformation condition to the hoisting controller;

[0008] The vibration detector is arranged on the hoisting arm and is used to detect the vibration condition of the hoisting arm when moving and send the detected vibration waveform to the hoisting controller;

[0009] The ultrasonic detector sends ultrasonic signals to the end surface of the steel cable connected with the goods on the hoisting arm, detects the stress condition of the steel cable in the working hoisting arm, and sends the detected ultrasonic echo to the hoisting controller;

[0010] The hoisting controller extracts the features of the deformation condition of the hoisting arm, the vibration waveform of the hoisting arm and the ultrasonic echo in the steel cable, and calculates the extracted features by using the calculation module to obtain the maximum speed of the hoisting motor, and the hoisting controller controls the hoisting motor to work below the maximum speed;

[0011] The hoisting controller sends the deformation condition of the hoisting arm, the vibration waveform of the hoisting arm and the ultrasonic echo in the steel cable to the upper computer for storage.

[0012] The stress sensor is arranged at different positions of the hoisting arm and is used to obtain the deformation amount of each position when the hoisting arm works, and a group of deformation amounts a1, a2, a3, …, an is obtained. n Sent to the hoisting controller, wherein n is the number of stress sensors, n≥5.

[0013] The stress sensors are uniformly arranged along the extension direction of the hoisting arm.

[0014] The vibration detector is an acceleration type vibration detector and can obtain the waveform of the vibration of the hoisting arm; the number of vibration detectors is the same as that of stress sensors, and the vibration detectors are installed close to the stress sensors;

[0015] The vibration detector obtains a group of vibration waveforms b1, b2, b3, …, bn and sends them to the hoisting controller, wherein n is the number of stress sensors, n≥5.

[0016] The ultrasonic detector is arranged on the end surface of the steel cable and includes an ultrasonic transmitter and an ultrasonic receiver; the ultrasonic wave emitted by the ultrasonic transmitter enters the steel cable from the end surface of the steel cable, propagates in the steel cable for one cycle, and is then output from the other end of the steel cable and received by the ultrasonic receiver;

[0017] The ultrasonic signal received by the receiver and the ultrasonic signal emitted by the ultrasonic emitter are sent to the hoisting controller together; the hoisting controller sends the ultrasonic emission signal and the ultrasonic receiving signal to the calculation module for calculation.

[0018] The calculation module obtains deformation variables a1, a2, a3, …, a n , a set of vibration waveforms b1, b2, b3, …, b n , and performs data processing on the ultrasonic emission signal and the ultrasonic receiving signal, and the specific manner is:

[0019] The calculation module compares the deformation variables and obtains the maximum value A of the deformation variables, wherein A = MAX[a1, a2, a3, …, a n ], that is, A is equal to the maximum value in a1, a2, a3, …, a n ; then a1, a2, a3, …, a n are calculated, s i = a i / A, wherein i = 1…n; s1, s2, s3, …, s n are obtained.

[0020] The calculation module performs noise reduction processing on the vibration waveforms, and the processed vibration waveforms are b'1, b'2, b'3, …, b' n ; the vibration waveforms after noise reduction are processed to obtain their frequency spectrum, and the frequency spectrum is superimposed to obtain the superimposed frequency spectrum B.

[0021] The calculation module calculates the time difference t of the ultrasonic emission signal and the ultrasonic receiving signal, that is, the propagation time t of the sound wave in the steel cable is obtained, and then the calculation module calculates the ultrasonic receiving signal to obtain the frequency spectrum S of the ultrasonic emission signal and the frequency spectrum T of the ultrasonic receiving signal, and calculates M = S-T, that is, the ultrasonic absorption spectrum M is obtained.

[0022] The calculation module inputs s1, s2, s3, …, s n , B, and M into the deep neural network model as inputs to obtain the full load coefficient of the hoisting arm, and the full load coefficient is the ratio of the current load of the hoisting arm to the full load of the hoisting arm G.

[0023] Then the maximum speed R of the hoisting motor is k·G·A; wherein k is an artificial set working coefficient.

[0024] The method for establishing the deep neural network is:

[0025] Load the load on the hoisting arm and keep the hoisting arm still, at this time, the stress sensor, the vibration detector, and the ultrasonic detector detect the deformation variables a1, a2, a3, …, a n , a set of vibration waveforms b1, b2, b3, …, b nand ultrasonic emission signal and ultrasonic receiving signal data;

[0026] Load different loads on the lifting arm, and calculate the deformation a1, a2, a3, …, a n a set of vibration waveforms b1, b2, b3, …, b n and ultrasonic emission signal and ultrasonic receiving signal data until the lifting arm reaches full load;

[0027] Thus, the deformation a1, a2, a3, …, a n a set of vibration waveforms b1, b2, b3, …, b n and ultrasonic emission signal and ultrasonic receiving signal data;

[0028] a set of deformation a1, a2, a3, …, a n a set of vibration waveforms b1, b2, b3, …, b n and ultrasonic emission signal and ultrasonic receiving signal data as input, and the full load coefficient as output to train the deep neural network model, so as to obtain a deep neural network model that can calculate the full load coefficient.

[0029] The stress sensor can detect the actual stress of the lifting arm during operation, the vibration waveform detected by the vibration detector reflects the load condition on one hand, and also reflects whether the lifting arm has various resonances during operation on the other hand; and the ultrasonic sensor can obtain the tension degree of the steel cable during operation, so that the lifting process can be comprehensively detected.

[0030] Since the deformation a1, a2, a3, …, a n a set of vibration waveforms b1, b2, b3, …, b n and ultrasonic emission signal and ultrasonic receiving signal are stable, the deformation a1, a2, a3, …, a n a set of vibration waveforms b1, b2, b3, …, b n and ultrasonic emission signal and ultrasonic receiving signal are changed due to the continuous loading of external driving, that is, the deformation a1, a2, a3, …, a n a set of vibration waveforms b1, b2, b3, …, b n and ultrasonic emission signal and ultrasonic receiving signal are continuously changing, and the deep neural network model established at the time of static can obtain the actual load of the lifting arm during movement, and the maximum operating speed of the lifting motor is obtained through the actual load, so as to ensure the temperature and safety of operation.

[0031] In addition, after the load coefficient is obtained, the stress of the lifting arm is constantly changing due to the constantly changing center of gravity during the lifting process, and the maximum deformation of the lifting arm is taken as a parameter to calculate the actual maximum operating speed, so that safety, stability and efficiency are more considered. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 It is a schematic diagram of the overall architecture of the present application. DETAILED DESCRIPTION

[0033] The advantages, features and methods of achieving the purposes of the present application will be clear through the drawings and the following detailed description.

[0034] Example 1:

[0035] A tower crane lifting mechanism embedded integrated control system, comprising a lifting controller, a calculation module, a lifting motor, a stress sensor, a vibration detector, an ultrasonic detector and an upper computer;

[0036] The lifting controller is embedded in the integrated architecture design, the calculation module, the lifting motor, the stress sensor, the ultrasonic detector and the vibration detector are all connected to the lifting controller, the lifting motor is used to drive the lifting module to rise and fall, thereby realizing the lifting work of the crane; the stress sensor is arranged on the lifting arm for detecting the deformation condition of the lifting arm when carrying goods, and sending the deformation condition to the lifting controller;

[0037] The vibration detector is arranged on the lifting arm for detecting the vibration condition of the lifting arm when moving, and sending the detected vibration waveform to the lifting controller;

[0038] The ultrasonic detector sends ultrasonic signals to the end surface of the steel cable connected to the goods on the lifting arm, detects the stress condition of the steel cable during the operation of the lifting arm, and sends the detected ultrasonic echo to the lifting controller;

[0039] The lifting controller extracts the features of the deformation condition of the lifting arm, the vibration waveform of the lifting arm and the ultrasonic echo in the steel cable, and calculates the extracted features by using the calculation module to obtain the maximum speed of the lifting motor, and the lifting controller controls the lifting motor to work below the maximum speed;

[0040] The lifting controller also sends the deformation condition of the lifting arm, the vibration waveform of the lifting arm and the ultrasonic echo in the steel cable to the upper computer for storage.

[0041] The stress sensor is a plurality of sensors arranged at different positions of the lifting arm for obtaining the deformation amount of each position when the lifting arm works, and obtaining a group of deformation amounts a1, a2, a3, …, an. n The deformation amounts are sent to the lifting controller, where n is the number of stress sensors, n≥5.

[0042] The stress sensors are uniformly arranged along the extension direction of the hoisting arm.

[0043] The vibration detector is an acceleration type vibration detector capable of acquiring a waveform of the vibration of the hoisting arm; the number of the vibration detectors is the same as that of the stress sensors, and the vibration detectors are installed close to the stress sensors;

[0044] The vibration detector acquires a group of vibration waveforms b1, b2, b3, …, bn and sends them to the hoisting controller, where n is the number of the stress sensors, and n≥5.

[0045] The ultrasonic detector is arranged on the end surface of the steel cable and includes an ultrasonic transmitter and an ultrasonic receiver; the ultrasonic wave emitted by the ultrasonic transmitter enters the steel cable from the end surface of the steel cable, propagates in the steel cable for one round, and is output from the other end of the steel cable and received by the ultrasonic receiver;

[0046] The ultrasonic wave signal received by the receiver and the ultrasonic signal emitted by the ultrasonic transmitter are sent to the hoisting controller together; the hoisting controller sends the ultrasonic emission signal and the ultrasonic reception signal to the calculation module for calculation.

[0047] The calculation module acquires deformation variables a1, a2, a3, …, a n , a group of vibration waveforms b1, b2, b3, …, b n , and the ultrasonic emission signal and the ultrasonic reception signal for data processing, and the specific way is:

[0048] The calculation module compares the deformation variables and acquires a maximum value A of the deformation variables, where A=MAX[a1, a2, a3, …, a n ], that is, A is equal to the maximum value in a1, a2, a3, …, a n ; then a1, a2, a3, …, a n are calculated, s i =a i / A, where i=1…n; s1, s2, s3, …, s n are obtained.

[0049] The calculation module performs noise reduction processing on the vibration waveforms, and the processed vibration waveforms are b’1, b’2, b’3, …, b’ n ; the vibration waveforms after noise reduction are processed to obtain their frequency spectrum, and the frequency spectrum is superimposed to obtain a superimposed frequency spectrum B.

[0050] The calculation module calculates the time difference t of the ultrasonic emission signal and the ultrasonic reception signal, that is, the propagation time t of the sound wave in the steel cable is obtained, and then the calculation module calculates the ultrasonic reception signal to obtain the frequency spectrum S of the ultrasonic emission signal and the frequency spectrum T of the ultrasonic reception signal, and calculates M=S-T, that is, the ultrasonic absorption spectrum M is obtained.

[0051] The computing module inputs s1, s2, s3,..., s n , B, M as inputs into the deep neural network model to obtain a full load coefficient of the lifting arm, the full load coefficient being a ratio of a current load of the lifting arm to a full load of the lifting arm G.

[0052] The maximum rotating speed R of the lifting motor is k G A, wherein k is a working coefficient set by a human being.

[0053] The method for establishing the deep neural network is as follows:

[0054] The lifting arm is loaded with a load, and the lifting arm is kept stationary, at which time the deformation variables a1, a2, a3,..., a n , a set of vibration waveforms b1, b2, b3,..., b n , and ultrasonic emission signal and ultrasonic receiving signal data are detected by the stress sensor, the vibration detector, and the ultrasonic detector.

[0055] Different loads are loaded on the lifting arm, and the deformation variables a1, a2, a3,..., a n , a set of vibration waveforms b1, b2, b3,..., b n , and ultrasonic emission signal and ultrasonic receiving signal data after each load are calculated until the lifting arm reaches full load.

[0056] Thus, the deformation variables a1, a2, a3,..., a n , a set of vibration waveforms b1, b2, b3,..., b n , and ultrasonic emission signal and ultrasonic receiving signal data corresponding to the full load coefficient are obtained.

[0057] The deformation variables a1, a2, a3,..., a n , a set of vibration waveforms b1, b2, b3,..., b n , and ultrasonic emission signal and ultrasonic receiving signal data are inputted, and the deep neural network model is trained with the full load coefficient as output, so that the deep neural network model for calculating the full load coefficient is obtained.

[0058] Embodiment 2:

[0059] This embodiment further introduces the specific working method of the system.

[0060] First, the deep neural network model is established,

[0061] Different loads are loaded on the lifting arm, and the deformation variables a1, a2, a3,..., a n , a set of vibration waveforms b1, b2, b3,..., b nand ultrasonic emission signal and ultrasonic receiving signal data until the lifting arm reaches full load;

[0062] Thus, the deformation variables a1, a2, a3, …, a n a set of vibration waveforms b1, b2, b3, …, b n and ultrasonic emission signal and ultrasonic receiving signal data;

[0063] in the form of deformation variables a1, a2, a3, …, a n a set of vibration waveforms b1, b2, b3, …, b n and ultrasonic emission signal and ultrasonic receiving signal data as input, the full load coefficient as output to train the deep neural network model, so as to obtain a deep neural network model capable of calculating the full load coefficient.

[0064] The number n of sensors in modeling can be selected as 10 or a larger value, so as to obtain the most comprehensive information, and in actual detection, the number of sensors can be slightly reduced.

[0065] Then, the data acquisition during operation is performed.

[0066] The stress sensors acquire the deformation variables of each position under stress during the operation of the lifting arm, and obtain a set of deformation variables a1, a2, a3, …, a n are sent to the lifting controller, where n is the number of stress sensors, n≥5. The stress sensors are uniformly arranged along the extension direction of the lifting arm.

[0067] The number of vibration detectors is the same as that of stress sensors, and the vibration detectors are installed close to the stress sensors; the vibration detectors acquire the waveforms of the vibration of the lifting arm;

[0068] The vibration detectors acquire a set of vibration waveforms b1, b2, b3, …, bn and send them to the lifting controller, where n is the number of stress sensors, n≥5.

[0069] The ultrasonic detectors control the ultrasonic waves emitted by the ultrasonic emitter to enter the steel cable from the end face of the steel cable, propagate around in the steel cable, and then output from the other end of the steel cable and be received by the ultrasonic receiver; the ultrasonic wave signals received by the receiver and the ultrasonic signals emitted by the ultrasonic emitter are sent to the lifting controller together; the lifting controller sends the ultrasonic emission signal and the ultrasonic receiving signal to the calculation module for calculation.

[0070] The calculation module acquires deformation variables a1, a2, a3, …, a n a set of vibration waveforms b1, b2, b3, …, b n and ultrasonic emission signal and ultrasonic receiving signal data for data processing, and the specific method is:

[0071] The calculation module compares the deformation variables and obtains the maximum value A of the deformation variables, where A = MAX[a1, a2, a3, ..., a n That is, A equals a1, a2, a3, ..., a n The maximum value in; then a1, a2, a3, ..., a n Perform calculations, s i =a i / A, where i=1…n; to obtain s1,s2,s3,…,s n ;

[0072] The calculation module performs noise reduction processing on the vibration waveform, resulting in vibration waveforms of type b'1, b'2, b'3, ..., b' n The noise-reduced vibration waveform is processed to obtain its spectrum, and the spectra are superimposed to obtain the superimposed spectrum B.

[0073] The calculation module calculates the time difference t between the ultrasonic transmitted signal and the ultrasonic received signal, thus obtaining the propagation time t of the sound wave in the steel cable. Then, the calculation module calculates the ultrasonic received signal to obtain the spectrum S of the ultrasonic transmitted signal and the spectrum T of the ultrasonic received signal. The calculation M=ST is then performed to obtain the ultrasonic absorption spectrum M.

[0074] The calculation module will calculate s1, s2, s3, ..., s n B and M are used as inputs to the deep neural network model to obtain the full load coefficient of the boom, which is the ratio of the current load of the boom to the full load of the boom, G.

[0075] The maximum speed of the hoisting motor is R = k·G·A; where k is a manually set working coefficient.

[0076] This invention designs a stress sensor, an ultrasonic detector, and a vibration detector to monitor the lifting process in real time. The stress sensor can detect the actual stress on the lifting arm during operation; the vibration detector detects the vibration waveform, which reflects both the load and whether there are various resonances in the lifting arm during operation; and the ultrasonic sensor can obtain the tension of the steel cable during operation, thus comprehensively monitoring the lifting process.

[0077] Since the deformation of the boom is a1, a2, a3, ..., a when it is stationary, n A set of vibration waveforms b1, b2, b3, ..., b n Furthermore, the ultrasonic transmission and reception signals are stable, and the deformations a1, a2, a3, ..., a2 during actual operation are stable. n A set of vibration waveforms b1, b2, b3, ..., b nAnd the ultrasonic emission signal and the ultrasonic receiving signal are changed due to the continuously loaded external drive, that is, in actual work, the deformation variables a1, a2, a3, …, a n , a set of vibration waveforms b1, b2, b3, …, b n And the ultrasonic emission signal and the ultrasonic receiving signal are continuously changed, the actual load of the lifting arm during movement can be obtained through the deep neural network model established at rest, and the maximum operating speed of the lifting motor is obtained through the actual load, which ensures the temperature and safety of operation.

[0078] In addition, after obtaining the load coefficient, due to the continuous change of the center of gravity during lifting, the stress of the lifting arm is also continuously changed, and the maximum deformation of the lifting arm is calculated as a parameter when calculating the actual maximum operating speed, which more greatly takes into account safety, stability and efficiency.

[0079] The above is only a preferred embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A tower crane hoisting mechanism embedded integrated control system, comprising a hoisting controller, a computing module, a hoisting motor, a stress sensor, a vibration detector, an ultrasonic detector and an upper computer; characterized in that: the hoisting controller is embedded in an integrated architecture design, the computing module, the hoisting motor, the stress sensor and the vibration detector are all connected to the hoisting controller, the hoisting motor is used to drive the hoisting module to rise and fall, thereby realizing the hoisting work of the crane; the stress sensor is arranged on the hoisting arm and is used to detect the deformation condition of the hoisting arm when it is moving with the load, and sends the deformation condition to the hoisting controller; the vibration detector is arranged on the hoisting arm and is used to detect the vibration condition of the hoisting arm when it is moving, and sends the detected vibration waveform to the hoisting controller; the vibration detector is an acceleration type vibration detector and can obtain the waveform of the vibration of the hoisting arm; the number of vibration detectors is the same as that of stress sensors, and the vibration detectors are installed close to the stress sensors; the vibration detector obtains a group of vibration waveforms b1, b2, b3, …, bn and sends them to the hoisting controller, where n is the number of vibration detectors, n≥5; the ultrasonic detector sends ultrasonic signals to the end face of the cable connected with the load on the hoisting arm, detects the stress condition of the cable in operation, and sends the detected ultrasonic echo to the hoisting controller; The stress sensors are multiple, arranged at different positions of the lifting arm, used for obtaining deformation amounts of stress occurrence at each position when the lifting arm works, to obtain a set of deformation amounts a1, a2, a3, …, a n are sent to the lifting controller, wherein n is the number of stress sensors, n≥5. the ultrasonic detector is arranged on the end face of the cable and comprises an ultrasonic transmitter and an ultrasonic receiver, the ultrasonic wave emitted by the ultrasonic transmitter enters the cable from the end face of the cable, propagates in the cable for one round and is then output from the other end of the cable and received by the ultrasonic receiver; the ultrasonic wave signal received by the receiver and the ultrasonic signal emitted by the ultrasonic transmitter are sent to the hoisting controller together; the hoisting controller sends the ultrasonic emission signal and the ultrasonic reception signal to the computing module for calculation; the hoisting controller extracts features from the deformation condition of the hoisting arm, the vibration waveform of the hoisting arm and the ultrasonic echo in the cable, calculates the extracted features by using the computing module, obtains the maximum speed of the hoisting motor, and controls the hoisting motor to work below the maximum speed; the hoisting controller also sends the deformation condition of the hoisting arm, the vibration waveform of the hoisting arm and the ultrasonic echo in the cable to the upper computer for storage; the computing module calculates the time difference t of the ultrasonic emission signal and the ultrasonic reception signal, i.e. obtains the propagation time t of the sound wave in the cable, then calculates the ultrasonic emission signal and the ultrasonic reception signal to obtain the frequency spectrum S of the ultrasonic emission signal and the frequency spectrum T of the ultrasonic reception signal, calculates M=S-T, i.e. obtains the ultrasonic absorption spectrum M; The computing module obtains deformation variables a1, a2, a3, …, a n A set of vibration waveforms b1, b2, b3, …, b n and the ultrasonic transmitting signal and the ultrasonic receiving signal are subjected to data processing, and the specific mode is: The calculation module compares the deformation variables and obtains the maximum value A of the deformation variables, where A = MAX[a1, a2, a3, ..., a n That is, A equals a1, a2, a3, ..., a n The maximum value in; then a1, a2, a3, ..., a n Perform calculations, s i =a i / A, where i=1…n; to obtain s1,s2,s3,…,s n ; The computing module denoises the vibration waveform, and the denoised vibration waveform is b'1, b'2, b'3,... n The denoised vibration waveform is processed to obtain a frequency spectrum, and the frequency spectra are superimposed to obtain a superimposed frequency spectrum B. then the maximum speed R of the hoisting motor is k·G·A; where k is the working coefficient set artificially; The computing module takes s1, s2, s3, …, s n , B, M as inputs, inputs into a deep neural network model, and obtains a full load coefficient of the lifting arm, that is, a load ratio G of the current load of the lifting arm to the full load of the lifting arm. the establishment method of the deep neural network is:

2. The tower crane hoisting mechanism embedded integrated control system according to claim 1, characterized in that: Load a load on the hoisting arm, and keep the hoisting arm still, at this time, the deformation variables a1, a2, a3, …, a are detected through the stress sensor, the vibration detector, and the ultrasonic detector n a set of vibration waveforms b1, b2, b3, …, b n and ultrasonic emission signal and ultrasonic receiving signal data; Loading different loads on the hoisting arm and calculating the deformation amount a1, a2, a3,..., a after each load is loaded n a set of vibration waveforms b1, b2, b3,... n and ultrasonic transmission signal and ultrasonic reception signal data until the hoisting arm reaches full load; From this the deformation variables a1, a2, a3,..., a corresponding to the load factors are obtained n a set of vibration waveforms b1, b2, b3,..., b n and ultrasound transmission and ultrasound reception signal data; with deformation variables a1, a2, a3,... n a set of vibration waveforms b1, b2, b3,... n and the ultrasound transmit signal and ultrasound receive signal data as input, the full load factor as output, a deep neural network model is trained, thereby obtaining a deep neural network model that is capable of calculating the full load factor. the stress sensors are uniformly arranged along the extension direction of the hoisting arm. ​

Images