Port portal crane point inspection management system

Abstract

The invention relates to a port portal crane point inspection management system. The invention relates to the technical field of port portal crane point inspection management, a magnetic flux balance principle is applied to a vulnerable section of a steel wire rope, a flaw detection sensor is added, and a steel wire rope operation condition online monitoring system is formed; an existing equipment lubricating system is upgraded, electromagnetic valves and flow meters are additionally arranged on lubricating pipelines in a parallel system, and a circulating lubricating system with the fixed point position, the fixed period and the fixed oil injection amount is achieved. On the aspects of real-time point inspection and abnormal state instant alarm, detection is planned to be carried out on key parts of a port portal crane and parts which are difficult to reach by daily point inspection, and research is planned to be carried out on two aspects of special point inspection equipment and a detection algorithm.

Description

Technical Field

[0001] The invention relates to the technical field of port gantry crane inspection management, in particular to a port gantry crane inspection management system. Background Art

[0002] Qinhuangdao Port Group Co., Ltd. proposed a monitoring method based on vibration sensors for gantry crane slewing bearings. The Fourier transform combined with wavelet transform method was used to complete the signal processing in the time domain and frequency domain during the working process to obtain the vibration characteristics under normal and abnormal conditions. Tianjin Port Authority established a set of vibration signal processing algorithms for the vibration of the motor and reducer system by establishing the dynamic model of the reducer and drum bearing. The vibration detection process was completed by reasonably selecting the observation points and feature vectors.

[0003] The existing fault detection methods have single functions and simple detection methods, and cannot effectively realize multi-dimensional real-time online detection of port gantry cranes. Traditional detection methods, the system mostly uses manual acquisition methods, which cannot achieve timely alarm.

[0004] In response to business pain points, real-time detection of key parts of port gantry cranes and parts that are extremely difficult to reach during daily inspections is achieved. On the one hand, the detection risk and fatigue of inspection personnel are greatly reduced, and on the other hand, safety hazards are eliminated to the greatest extent to achieve continuous monitoring; the special computing visual feature fusion, optical analysis and array eddy current combination for gantry cranes proposed in this technical solution will significantly improve the sensitivity and efficiency of damage monitoring; remote mobile inspection, unlike traditional video surveillance, is most convenient for users to use, which is a subversion of traditional monitoring and inspection methods, and provides a solid technical foundation for the finalization and promotion of future products; the proposed anti-interference methods such as harsh environment, mechanism motion interference, and camera self-stabilization will play an updating and leading role in the current industry technology. Summary of the invention

[0005] In view of the shortcomings of the existing technology, the magnetic flux balance principle is applied to the vulnerable section of the wire rope, and a flaw detection sensor is added to form an online monitoring system for the operation status of the wire rope; the lubrication system of the existing equipment is upgraded, and solenoid valves and flow meters are added to each lubrication pipeline in the parallel system to realize a circulating lubrication system with fixed point position, fixed cycle and fixed oil injection amount. In terms of real-time inspection and instant alarm of abnormal status, it is planned to detect the key parts of the port gantry crane and the parts that are extremely difficult to reach during daily inspection, and it is planned to complete research on special inspection equipment and detection algorithms. The present invention provides a port gantry crane inspection management system.

[0006] The present invention provides the following technical solutions:

[0007] A port gantry crane inspection management system, the system comprising:

[0008] Vibration detection system, which performs vibration detection through bearing vibration detection sensor collection equipment;

[0009] The wire rope flaw detection system can realize the comprehensive detection of wire rope broken wires, wear, corrosion, deformation, loose strands and other defects through the wire rope flaw detection sensor collection equipment, and accurately measure the internal broken wires;

[0010] Automatic lubrication system, automatic lubrication through automatic lubrication system quantitative distribution valve and controller;

[0011] The visual positioning system for key parts of the wire rope performs flaw detection through visual system laser and camera equipment.

[0012] Preferably, the vibration detection system is directly connected to the corresponding control device through an integrated acceleration vibration sensor to achieve online monitoring, and with the help of a bus integrated communication system, an effective analysis of the vibration signal under motion interference is achieved based on high-order signal means;

[0013] The integrated acceleration vibration sensor integrates the acceleration vibration measuring element and the transmitter module into one, and can directly output the acceleration value signal linearly. The motor equipment status monitoring method based on wavelet analysis and neural network uses wavelet transform to extract the vibration signal characteristics, uses machine learning to identify the characteristics, and outputs the equipment status; uses wavelet time-frequency analysis technology to eliminate noise and filter the motor vibration signal, obtains the frequency band energy through wavelet packet decomposition coefficients, and extracts fault characteristics according to the changes in the energy of each frequency band;

[0014] Collect vibration sample signals and test signals respectively, use wavelet packets to perform offline analysis and feature extraction on the signals, and obtain sample signal feature vectors and test signal feature vectors respectively; use machine learning or threshold methods to determine the device status.

[0015] Preferably, the wire rope flaw detection system uses multiple groups of flaw detection sensor devices to respectively collect the time domain signals of the wire rope magnetic flux at distributed key points, and uses the magnetic flux analysis method and accumulated data knowledge to identify the state of the wire rope to obtain a comprehensive damage image of the wire rope.

[0016] Preferably, the automatic lubrication system completes the interactive control process of lubrication through the oil supply controller PLC and the upper industrial computer via TCP / IP. The lubrication plan is manually formulated in the upper controller. When the lubrication plan is completed and released, the upper controller sends the lubrication time and lubrication amount to the PLC at regular intervals. The PLC receives the instructions to complete the closed-loop control process.

[0017] Preferably, the visual positioning system for key parts of the wire rope adopts two detection modules, each module contains a specially developed optical spot inspection function module for online monitoring of the operation status of the wire rope. The online monitoring optical spot inspection function module can resist extreme sunlight conditions and working vibration interference, and adopts precise optical circuits and precisely developed visual measurement technology to study the changes in the optical properties of the detection object; the array eddy current uses the array eddy current sensor to establish a three-dimensional model for hidden defect detection, optimizes the detection signal space between the center of the detection coil and the detection object, studies the attenuation trend of the defect signal in the weak magnetic field area between adjacent channels of the array, and adopts an array-placed detection method to realize the crystal oscillator acquisition of the detection signal characteristics of defects of different sizes.

[0018] A method for detecting a wire rope state based on vision, the method comprising the following steps:

[0019] Collect the original wire rope image and pre-process the original image;

[0020] Perform feature analysis on the preprocessed image to extract the features of the target;

[0021] According to the extracted target features, conditions are set to perform positioning and settlement. After the original image is processed through a series of algorithms such as preprocessing, feature analysis, spatial positioning, and key spatial response, the spatial position of the wire rope and pulley is obtained. Finally, based on the spatial positioning, the normal and abnormal data of the state are directly obtained through logical reasoning.

[0022] Preferably, the image processing process mainly includes two steps: binocular correction and edge detection. The principle of image correction is to reproject the image planes in two different directions to the same plane through matrix transformation, with the optical axes parallel to each other, under the premise of ensuring that the relative positions of the optical centers of the two cameras remain unchanged. In this way, the epipolar lines of the two cameras are horizontally aligned, and when matching, it is only necessary to search for corresponding points along the horizontal epipolar line direction.

[0023] When the wire rope features are not obvious, edge detection is used to make the rope edge contour obvious and reduce the feature matching error rate; after binocular correction and edge detection, the left and right image polar lines are horizontally aligned, the rope edge features in the image are more obvious, and then feature matching is performed on the left and right images.

[0024] Preferably, feature matching is divided into three steps:

[0025] 1. Feature extraction, that is, extracting key points in the image as feature points of the image;

[0026] 2. Feature description, that is, using vectors to describe the image information of feature points, requiring that the feature vectors correspond to the feature points one by one;

[0027] 3. Feature matching, that is, calculating the distance between feature vectors.

[0028] A computer-readable storage medium stores a computer program, which is executed by a processor to implement a vision-based wire rope state detection method.

[0029] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements a vision-based wire rope state detection method when executing the computer program.

[0030] The present invention has the following beneficial effects:

[0031] Compared with the prior art, the present invention has the following advantages:

[0032] The present invention can realize real-time and continuous detection of key parts of port gantry cranes and parts that are extremely difficult to reach during daily inspections;

[0033] (2) Circulating lubrication system to improve the effect of lubrication parts;

[0034] (3) R&D technology can support equipment availability rate ≥ 98%;

[0035] (3) The operating reliability of system automation equipment is ≥ 98%;

[0036] (4) The system automation software operation reliability shall be ≥ 99%;

[0037] (5) System missed detection rate ≤ 2%;

[0038] (6) System false alarm ≤ 5%;

[0039] (7) The researched technology can be subsequently commercialized, and the cost of commercialization shall not change by more than 20%. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 The system architecture of the present invention;

[0042] Figure 2 It is the principle diagram of magnetic flux change flaw detection;

[0043] Figure 3 This is a schematic diagram of the installation of the flaw detection sensor;

[0044] Figure 4 Lubrication hardware architecture diagram

[0045] Figure 5 It is a flow chart of lubrication system;

[0046] Figure 6 It is the PLC control block diagram;

[0047] Figure 7 It’s the monitoring process. DETAILED DESCRIPTION

[0048] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0050] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0051] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0052] The present invention is described in detail below in conjunction with specific embodiments. Specific embodiment one:

[0054] according to Figure 1-Figure 7 As shown, the specific optimization technical solution adopted by the present invention to solve the above technical problems is: the present invention relates to a port gantry crane inspection management system.

[0055] The system comprises:

[0056] Vibration detection system, which performs vibration detection through bearing vibration detection sensor collection equipment;

[0057] The wire rope flaw detection system can realize the comprehensive detection of wire rope broken wires, wear, corrosion, deformation, loose strands and other defects through the wire rope flaw detection sensor collection equipment, and accurately measure the internal broken wires;

[0058] Automatic lubrication system, automatic lubrication through automatic lubrication system quantitative distribution valve and controller;

[0059] The visual positioning system for key parts of the wire rope performs flaw detection through visual system laser and camera equipment. Specific embodiment 2:

[0061] The difference between the second embodiment of the present invention and the first embodiment is that:

[0062] The vibration detection system is directly connected to the corresponding control equipment through the integrated acceleration vibration sensor to achieve online monitoring. With the help of the bus integrated communication system, the vibration signal under motion interference can be effectively analyzed based on high-order signal means.

[0063] The integrated acceleration vibration sensor integrates the acceleration vibration measuring element and the transmitter module into one, and can directly output the acceleration value signal linearly. The motor equipment status monitoring method based on wavelet analysis and neural network uses wavelet transform to extract the vibration signal characteristics, uses machine learning to identify the characteristics, and outputs the equipment status; uses wavelet time-frequency analysis technology to eliminate noise and filter the motor vibration signal, obtains the frequency band energy through wavelet packet decomposition coefficients, and extracts fault characteristics according to the changes in the energy of each frequency band;

[0064] Collect vibration sample signals and test signals respectively, use wavelet packets to perform offline analysis and feature extraction on the signals, and obtain sample signal feature vectors and test signal feature vectors respectively; use machine learning or threshold methods to determine the device status. Specific embodiment three:

[0066] The difference between the third embodiment of the present invention and the second embodiment is that:

[0067] The wire rope flaw detection system uses multiple groups of flaw detection sensor equipment to collect the time domain signals of the wire rope magnetic flux at distributed key points, and uses the magnetic flux analysis method and accumulated data knowledge to identify the state of the wire rope and obtain a comprehensive damage image of the wire rope. Specific embodiment four:

[0069] The difference between the fourth embodiment of the present invention and the third embodiment is that:

[0070] The automatic lubrication system completes the interactive control process of lubrication through the oil supply controller PLC and the upper industrial computer via TCP / IP. The lubrication plan is manually formulated in the upper controller. When the lubrication plan is completed and released, the upper controller sends the lubrication time and lubrication amount to the PLC at a regular interval. The PLC receives the instructions to complete the closed-loop control process. Specific embodiment five:

[0072] The difference between the fifth embodiment of the present invention and the fourth embodiment is that:

[0073] The visual positioning system for key parts of wire ropes adopts two detection modules, each of which contains a specially developed optical spot inspection function module for online monitoring of the operation status of wire ropes. The online monitoring optical spot inspection function module can resist extreme sunlight conditions and working vibration interference, and adopts precise optical circuits and accurately developed visual measurement technology to study the changes in the optical properties of the detection object. The array eddy current uses the array eddy current sensor to establish a three-dimensional model for hidden defect detection, optimizes the detection signal space between the center of the detection coil and the detection object, studies the defect signal attenuation trend in the weak magnetic field area between adjacent channels of the array, and adopts an array-placed detection method to realize the crystal oscillator acquisition of detection signal characteristics of defects of different sizes. Specific embodiment six:

[0075] The difference between the sixth embodiment of the present invention and the fifth embodiment is that:

[0076] The present invention provides a method for detecting the state of a steel wire rope based on vision, which is characterized in that the method comprises the following steps:

[0077] Collect the original wire rope image and pre-process the original image;

[0078] Perform feature analysis on the preprocessed image to extract the features of the target;

[0079] According to the extracted target features, conditions are set to perform positioning and settlement. After the original image is processed through a series of algorithms such as preprocessing, feature analysis, spatial positioning, and key spatial response, the spatial position of the wire rope and pulley is obtained. Finally, based on the spatial positioning, the normal and abnormal data of the state are directly obtained through logical reasoning. Specific embodiment seven:

[0081] The difference between the seventh embodiment of the present invention and the sixth embodiment is that:

[0082] The image processing process mainly includes two steps: binocular correction and edge detection. The principle of image correction is to reproject the image planes in two different directions to the same plane through matrix transformation, while ensuring that the relative positions of the optical centers of the two cameras remain unchanged, and the optical axes are parallel to each other. In this way, the epipolar lines of the two cameras are horizontally aligned, and matching only requires searching for corresponding points along the horizontal epipolar line direction.

[0083] When the wire rope features are not obvious, edge detection is used to make the rope edge contour obvious and reduce the feature matching error rate; after binocular correction and edge detection, the left and right image polar lines are horizontally aligned, the rope edge features in the image are more obvious, and then feature matching is performed on the left and right images. Specific embodiment eight:

[0085] The difference between the eighth embodiment of the present invention and the seventh embodiment is that:

[0086] Feature matching is divided into 3 steps:

[0087] 1. Feature extraction, that is, extracting key points in the image as feature points of the image;

[0088] 2. Feature description, that is, using vectors to describe the image information of feature points, requiring that the feature vectors correspond to the feature points one by one;

[0089] 3. Feature matching, that is, calculating the distance between feature vectors. Specific embodiment nine:

[0091] The difference between the ninth embodiment of the present invention and the eighth embodiment is that:

[0092] The present invention provides a computer-readable storage medium on which a computer program is stored. The program is executed by a processor to implement a vision-based wire rope state detection method. Specific embodiment ten:

[0094] The difference between the tenth embodiment of the present invention and the ninth embodiment is that:

[0095] The present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements a vision-based wire rope state detection method when executing the computer program. Specific implementation method eleven:

[0097] The port gantry crane inspection management system proposed in this invention integrates multi-dimensional measurement, integrates machine vision and multiple physical signal analysis to realize all-round monitoring of key parts of the port gantry crane. The integrated system designed in this study is a pioneering study and is the first in the industry. At present, the most commonly used method in port gantry cranes is a combination of automatic monitoring and manual inspection. The main monitoring contents of port gantry cranes include key parts such as metal structures, wire ropes and bearings.

[0098] The developed fixed-point detection module will be placed in the key parts of the crane and the parts that are extremely difficult to reach during daily inspections, forming a visual-based real-time inspection and abnormal state instant alarm system for the wire rope; the magnetic flux balance principle is applied to the vulnerable section of the wire rope, and flaw detection sensors are added to form an online monitoring system for the operation status of the wire rope; the lubrication system of the existing equipment is upgraded, and solenoid valves and flow meters are added to each lubrication pipeline in the parallel system to realize a circulating lubrication system with fixed points, fixed cycles and fixed oil injection. In terms of real-time inspection and abnormal state instant alarm, it is planned to detect the key parts of the port gantry crane and the parts that are extremely difficult to reach during daily inspections, and it is planned to complete research on special inspection equipment and detection algorithms.

[0099] At the hardware equipment level, the main focus is on solving problems such as harsh detection environment, mechanism motion interference, and camera self-stabilization during the detection process. We develop and design a new type of high-precision visual system to solve the problems of large detection range and detection bias due to a single method in principle. We design a stable platform sensor subsystem and study stabilization devices to ensure that there is no position disturbance that affects the detection accuracy during the operation of the gantry crane.

[0100] At the detection algorithm level, we study the structure, texture, edge, and feature detection technologies under extreme outdoor lighting conditions to achieve robust optical detection of special equipment. We study high-precision grating projection structured light technology to achieve structural reconstruction and tiny deformation detection under harsh ambient light conditions. We implement a robust four-step phase shift method to ensure detection effectiveness and real-time performance. We study array eddy current detection technology, establish a three-dimensional model for hidden defect detection, and use an array placement detection method to achieve crystal oscillator acquisition of detection signal characteristics for defects of different sizes.

[0101] The research of this invention can communicate with the entire system and design a special equipment detection APP to achieve remote supervision and intervention of real-time measurement and monitoring.

[0102] The hardware of the gantry crane inspection system includes: bearing vibration detection sensor collection equipment, wire rope flaw detection sensor collection equipment, automatic lubrication system quantitative distribution valve and controller, visual system laser and camera equipment, power supply system and operating room electrical cabinet, etc.

[0103] Based on the original mature technical solution for flaw detection of wire ropes for mining hoists, this solution supplements and optimizes the relevant contents in view of the movement characteristics and working environment of the wire ropes of gantry cranes and in combination with previous technical exchanges. The latest magnetic balance technology is adopted to realize the comprehensive detection of wire rope broken wires, wear, corrosion, deformation, loose strands and other defects; it can accurately measure internal broken wires.

[0104] like Figure 2As shown, the excitation source can be (the number of turns of coils A and B is the same), and the induced electromotive force generated in coil C is zero. Regardless of whether the wire rope is shaking or moving, the output of coil C is zero. When the wire rope is defective, as shown in the figure: it first passes through coil A, the magnetic flux of coil A will change, and the magnetic flux balance will be broken at this time, and an induced electromotive force +ε will be generated in coil C; when the defect passes through coil B, it causes the magnetic flux of coil B to change, and an induced electromotive force -ε is generated in coil C. Whether the internal or external defects of the wire rope will cause changes in magnetic flux, that is, whether the wire rope is broken inside or outside, the sensor can measure it and output positive and negative signals. Because it adopts the balance principle, the measurement sensitivity is very high. In addition, this principle can overcome the influence of wire rope shaking on damage measurement and realize the equivalent measurement of internal and external broken wires of wire rope.

[0105] Multiple groups of flaw detection sensor equipment collect the time domain signals of the magnetic flux of the wire rope at distributed key points respectively, and use the magnetic flux analysis method and accumulated data knowledge to identify the state of the wire rope to obtain the comprehensive damage image of the wire rope. The main function of the distributed sensor equipment is to collect signals and pre-process signals. The data transmission process between the sensor and the controller is completed by wire. Multiple groups of sensor equipment correspond to one group of central control terminals. The central control terminal completes the post-processing of the data and transmits the data to the centralized control system through TCP.

[0106] The wire rope detection sensor is installed on the trunk platform of the gantry crane, mainly monitoring the status of the wire rope on one side of the wire rope hook. The position of the wire rope detection sensor is shown in the figure. There are two solutions for installing one sensor and two sensors for one wire rope respectively.

[0107] The mechanical components of the wire rope online monitoring equipment are as follows:

[0108] (1) Welding base: Need to be welded to the base plate and fixed with bolts to facilitate the removal of the sensor, replacement of the wire rope and maintenance.

[0109] (2) Sensor fixing base: Connect the welding base and fix the sensor to ensure the stability of the sensor and prevent it from swinging. It is easy to disassemble.

[0110] (3) Sensors: including magnetic heads and magnetic flux sensors, which monitor the operation of the wire rope, collect data, and send the data to the computer for analysis.

[0111] The hardware of the wire rope flaw detection system includes multiple sets of magnetizers and magnetic flux sensors. The magnetizers are used to generate detection magnetic fields and magnetic flux loops in the wire ropes, and the magnetic flux sensors are used to detect changes in magnetic flux in the wire ropes. When defects occur in the wire ropes, the magnetic flux sensors can detect tiny changes in magnetic flux.

[0112] Circulating lubrication system based on PLC control

[0113] The oil supply controller PLC and the upper industrial computer complete the interactive control process of lubrication through TCP / IP. The lubrication plan is manually formulated in the upper controller. When the lubrication plan is completed and released, the upper controller sends the lubrication time and lubrication amount to the PLC at a fixed time, and the PLC receives the instructions to complete the closed-loop control process.

[0114] The logic control system consists of two parts: a control module and a communication module. Based on the original parallel lubrication system, this solution adds valves and metering methods to the main oil injection circuit, performs statistical analysis on the acquired data, optimizes the lubrication control logic, controls the quantitative valve through PLC, and feeds back the oil injection amount through the counter. The PLC interacts with the industrial computer for data exchange, thus realizing the closed-loop oil injection control of the entire lubrication system.

[0115] The control module runs in the PLC and its main function is to realize the quantitative and timed oil supply of the lubrication system. The communication module refers to the TCP communication between the PLC and the host computer to transmit the oil signal.

[0116] The table above shows the communication codes between the lubrication system and the electrical cabinet. The two complete the exchange of commands through TCP socket.

[0117] (2) Control module

[0118] The control system diagram of the oiling quantity in the PLC of the present invention, the PC sends two variables "Quantity_o_oil" and "State_o_oil" to the PLC according to the specified time and oiling plan, and the PLC forms a closed-loop control with the help of a counter according to the oiling quantity and state. Since a quantitative valve is used, the oiling quantity information of the oiling point can be calculated by counting the number of valve core changes.

[0119] The lubrication system is divided into two parts: automatic control and manual control. The automatic control system is determined by the upper computer software to determine the lubrication time and amount. When the system is in automatic mode, the operator formulates a lubrication plan through the upper computer interface, including the lubrication amount and lubrication interval. After the operator has formulated the lubrication plan, the current time is used as the start time, and the upper computer sends a lubrication instruction to the lower PLC every "set time" to complete the scheduled lubrication.

[0120] Taking into account the particularity that the lubrication site is closely related to the on-site equipment, the manual oiling switch is installed on the door machine next to the lubrication pump, and the number of oiling times is set directly by PLC.

[0121] For the existing lubrication system, the present invention adopts the form of a quantitative valve to complete the automatic control for the pipeline to be modified. Since the lubrication part is a modification of the existing project, the present invention replaces the valve block that cannot be measured on the existing gantry equipment with a valve block that can be measured. The pipeline uses the existing pipeline, so the installation position is the same as the existing installation position, as marked in the figure. By installing sensing and control equipment at the corresponding position, real-time monitoring of the terminal can be achieved.

[0122] The quantitative distribution valve is used to complete the equipment transformation in the existing pipeline, which is achieved by replacing the quantitative distribution valve body.

[0123] By controlling the two main lines to circulate and time-share the oil supply, the control piston and the quantitative piston can be moved in a circular manner, thereby realizing the quantitative oil supply process. By comprehensively designing the size of the quantitative piston and counting the number of oiling times, the oiling process can be monitored. The quantitative oiling solution based on the distribution valve adopts PLC control and uses the oiling times to realize the precise lubrication process of each lubrication point.

[0124] Lubrication system testing

[0125] The lubrication system is a modification of an existing project. The laboratory environment lacks lubrication pumps and lubrication pipelines, so it is impossible to perform simulation in the laboratory and only functional tests have been completed.

[0126] The lubrication system adds a counting function to each lubrication point, and completes the lubrication work of the lubrication point through the quantitative valve and the counting switch. The present invention performs a manual simulation action on the counting switch to determine that the lubrication system can count correctly. At the same time, a communication test is performed between the lubrication PLC and the host computer.

[0127] Vibration Detection System Based on Spectrum Analysis

[0128] Through the integrated acceleration vibration sensor, the corresponding control equipment can be directly connected to the inspection object on site to realize online monitoring. With the help of the bus integrated communication system, low-power consumption, low-cost, multi-node, multi-topology, low complexity, fast, reliable and safe sensor networking can be realized. At the same time, based on the computing platform, parallel and real-time analysis of multiple sensor signals can be realized. Based on high-order signal means, effective analysis of vibration signals under motion interference can be realized. With the help of spatial decoupling signal decoupling technology and blind source analysis technology, equipment status judgment and positioning can be realized. Continuous intelligent discrimination is realized for stable feature quantities based on a large amount of data.

[0129] Sensor Principle

[0130] The integrated acceleration vibration sensor integrates the acceleration vibration measuring element and the transmitter module, which can directly output the acceleration value signal linearly. It is designed with a convenient and stable installation mechanism for the inspection object, directly connected to the corresponding control equipment on site, and simplifies the commissioning. During the research and development process, the compatibility and promotion of the system are considered at the same time, making the system have the potential to be widely used in the real-time monitoring of bearing vibration of rotating machinery.

[0131] During the stable operation of the system, the equipment will produce gentle vibrations, and its spectrum curve is relatively smooth. When the equipment fails, the vibration will produce random and regular noise, which is specifically manifested as an increase in frequency in the frequency domain. By analyzing the first-order, second-order, and higher-order feature quantities of the time domain signal, multi-dimensional features in the time and frequency domain are obtained, and the equipment status is evaluated based on threshold methods or machine learning.

[0132] The motor equipment status monitoring method based on wavelet analysis and neural network uses wavelet transform to extract vibration signal features, uses machine learning to identify features, and outputs equipment status. Wavelet time-frequency analysis technology is used to denoise and filter the motor vibration signal, and the frequency band energy is obtained by wavelet packet decomposition coefficients, and fault features are extracted according to the changes in the energy of each frequency band. Vibration sample signals and test signals are collected separately, and wavelet packets are used to perform offline analysis and feature extraction on the signals to obtain sample signal feature vectors and test signal feature vectors respectively; machine learning or threshold methods are used to determine the equipment status.

[0133] In addition, the instantaneous frequency of the vibration signal can be estimated by using the peak value of the short-time Fourier transform spectrum based on the instantaneous frequency rotation tracking method, and the instantaneous frequency changes in the low-frequency band. The nonlinear short-time Fourier transform accurately estimates the instantaneous frequency of the vibration signal, integrates the instantaneous frequency curve to obtain the instantaneous phase curve, and completes the tracking process through resampling.

[0134] The judgment method and criteria are as follows: The general indicators selected are the root mean square of vibration velocity (vibration intensity), the root mean square of vibration acceleration in multiple frequency bands, peak value, optimized kurtosis, steady impact intensity and other indicators. The thresholds will be set in combination with relevant standards, previous equipment cases of the same type and similar power, and the equipment's own indicator trends.

[0135] For the lifting and closing motor, the luffing mechanism motor and the rotating spindle motor, it is planned to use the root mean square vibration velocity, the root mean square vibration acceleration of multiple frequency bands, the steady impact intensity and the optimized kurtosis index: the root mean square vibration velocity is set according to the reference standard ISO20816-3 based on the motor power, speed and installation method; the root mean square vibration acceleration of multiple frequency bands is set according to the relative standard and with reference to other cases with similar power and equipment working conditions; the steady impact intensity is set based on the improved indicators and experimental procedures of the standard ISO17359-3.

[0136] The variable amplitude mechanism reducer is planned to adopt the root mean square of vibration velocity, the root mean square of vibration acceleration in multiple frequency bands, the peak value and the optimized kurtosis index: the root mean square of vibration velocity and the peak value are set according to the reference standard ISO20816-9 based on the gear structure type, whether it is closed and the number of transmission stages; the root mean square of vibration acceleration in multiple frequency bands is set based on relative standards, reference to other structures and similar working conditions and on-site experimental measurements.

[0137] The rotating spindle bearing is planned to adopt the root mean square vibration velocity, root mean square vibration acceleration in multiple frequency bands, and peak indicators: the root mean square vibration velocity, peak value, root mean square vibration acceleration in multiple frequency bands are set based on relative standards, reference standard ISO10816-21 (for reference only), other cases with similar structures and working conditions, and on-site experimental measurements.

[0138] The hinge points of the luffing mechanism and boom system, the lower hinge point of the small pull rod, the upper hinge point of the small pull rod, the lower hinge point of the boom, the lower hinge point of the large pull rod, the upper hinge point of the large pull rod, the upper hinge point of the boom, and the hinge points of the balance system are planned to adopt the root mean square of vibration velocity, the root mean square of vibration acceleration in multiple frequency bands, and the peak index: the root mean square of vibration velocity, peak value, and root mean square of vibration acceleration in multiple frequency bands are set based on relative standards, reference to other structures and similar working conditions, and on-site experimental measurements.

[0139] Distributed sensor communication system

[0140] As shown in the communication network description, the system is implemented by wired network. The wired network uses analog or RS485 / RS232 communication. The advantage of analog is that the acquisition frequency is relatively high, but its signal is obviously affected by distance and external interference. The advantage of RS485 / RS232 communication is that the data transmission distance is long and the transmission process is reliable.

[0141] The present invention uses a vibration sensor to monitor the bearing status. The model is HITHMAIC-PSC-freq. The above table shows the installation position and quantity of the sensor. All sensor models are the same. The sensor adopts a welding base and a sensor magnetic suction method. The sensor base and the equipment are connected by welding at the above position. The sensor base uses magnetic materials to facilitate sensor suction and replacement of the sensor in the later stage.

[0142] Vibration system testing

[0143] Since the present invention is a research and development project rather than a supply project, the problem it solves is that there is no dedicated, ready-made equipment on the market. Therefore, it is impossible to simulate the on-site vibration conditions of the gantry crane bearing in the laboratory. The vibration environment cannot be simulated in this research laboratory, so this test only verifies the vibration function.

[0144] In order to test the basic functions of the vibration sensor, this study carried out a functional simulation experiment on the final electrical system. The system continuously collected data from all channels at multiple times to observe data quality and packet loss. The following figure shows the test process and sensor connection.

[0145] During the test, all sensors were kept stationary to check their dynamic measurement accuracy. The results are shown in the figure above. The dynamic measurement accuracy of the sensors meets the measurement requirements.

[0146] Real-time gantry crane wire rope inspection and abnormal status instant alarm based on vision

[0147] Vision system design

[0148] Taking the No. 3 gantry crane of Caofeidian Hegang Industrial Logistics Port as an example, the key parts of the crane and the parts that are extremely difficult to reach during daily inspections are placed on the fixed-point detection module as shown in the figure:

[0149] Each group uses two detection modules, and each module contains a directional research and development of the wire rope running condition online monitoring optical spot inspection function module. The online monitoring optical spot inspection function module can resist extreme sunlight and working vibration interference, and adopts precision optical circuits and precisely developed visual measurement technology to study the changes in the optical characteristics of the detection object; the array eddy current uses the array eddy current sensor to establish a three-dimensional model for hidden defect detection, optimize the detection coil center and the detection object detection signal space, study the defect signal attenuation trend in the weak magnetic field area between adjacent channels of the array, and adopt an array placement detection method to achieve the crystal oscillator acquisition of the detection signal characteristics of defects of different sizes.

[0150] Optical inspection of key parts of port gantry cranes and parts that are extremely difficult to reach during daily inspections has problems such as harsh inspection environment and interference from mechanism movement. Therefore, a robust optical inspection subsystem is designed as shown in the figure. The optical element surface quality inspection lens includes: a light source, a beam splitter unit, a focusing unit, and a detection unit; the first beam splitter unit is used to split the light emitted by the light source into two parts: detection light and reference light, and the two parts of light are incident on the first focusing unit and the second focusing unit respectively; the first focusing unit is used to receive the detection light, so that it becomes convergent light after passing through and is focused on the measurement. The second beam splitter unit is used to receive the detection light reflected from the measurement and passing through the first focusing unit and the reference light passing through the second focusing unit, and analyze the error information of the measurement object by measuring statistical characteristics.

[0151] The fixed point-line measurement system is fixed on one side of the steel belt for measurement. It includes a sliding bracket on which the point measurement units are deployed. The horizontal deployment position can be adjusted according to the test needs. The connecting bracket of the fixed bracket is bolted to the fixed bracket and the height is adjustable, which is convenient for later maintenance and upgrades.

[0152] The figure above shows the principle diagram of wire rope detection based on vision. The pulley is illuminated by a laser module to enhance the effect of the wire rope in the pulley. In the enhanced image, a denoising algorithm is introduced to remove image noise caused by sunlight. The denoised image can highlight the relative relationship between the wire rope and the edge of the pulley. The feature matrix of the image is calculated using the edge detection algorithm and the gradient algorithm. The feature matching and screening process is completed using a machine learning algorithm.

[0153] After the original image is processed through a series of algorithms such as preprocessing, feature analysis, spatial positioning, and key spatial response, the spatial position of the wire rope and pulley is obtained. Finally, based on the spatial positioning, the normal and abnormal data of the state are directly obtained through logical reasoning.

[0154] The private visual measurement system is mainly composed of image acquisition module, image processing module, and other software and hardware collaborative interaction modules (control module). After the collected image data is input into the computer, the image processing module completes the processing, extracts the characteristics of the target, such as quantity, position, length, etc., and then locates and solves it according to the preset condition values.

[0155] Binocular vision measurement is based on the parallax principle. As shown in the figure, the optical axes of the two cameras are parallel and perpendicular to the baseline of the binocular stereo vision model. The baseline distance B is the distance between the projection centers of the two cameras, and the focal length of the camera is f.

[0156] Assume that the left and right cameras simultaneously observe the same feature point P (xc, yc, zc) of the spatial object and obtain the image of point P, whose image coordinates are Pl = (x1, y1) and Pr = (x2, y2). Now the images of the two cameras are on the same plane, so the image coordinates Y of the feature point P are the same, that is, yl = yr = y, then from the triangular geometric relationship, we can get:

[0157]

[0158] Parallax is: D = x1-x2

[0159] It can be seen that the three-dimensional coordinates of the feature point P in the camera coordinate system are:

[0160]

[0161] Therefore, any point on the left camera image can determine the three-dimensional coordinates of the point as long as the corresponding matching point can be found on the right camera image. By deducing the triangular geometric relationship, we can get:

[0162] D=Bf / zc

[0163] When the baseline B and focal length f are fixed, that is, the parameters and relative position and posture of the camera are fixed, the parallax is inversely proportional to the depth of the spatial point. In order to quickly and accurately match the left camera image (hereinafter referred to as the left image) and the right camera image (hereinafter referred to as the right image), the left and right images need to be processed. The image processing process mainly includes two major links: binocular correction and edge detection. It is difficult to ensure that the left and right cameras are coplanar and the optical axes are parallel during actual shooting, which may cause image distortion. The image is not horizontally aligned, and the traditional feature matching strategy is a two-dimensional search, that is, traversing the entire image, which greatly reduces the accuracy and efficiency of matching. The principle of image correction is to reproject the image planes in two different directions to the same plane through matrix transformation and the optical axes are parallel to each other, while ensuring that the relative positions of the optical centers of the two cameras are fixed. In this way, the epipolar lines of the two cameras are horizontally aligned, and when matching, it is only necessary to search for corresponding points along the horizontal epipolar line direction. If the wire rope features are not obvious, edge detection can be used to make the rope edge contour obvious and reduce the feature matching error rate. Common edge detection methods include differential operators, Canny operators, and Laplacian of Gaussian (LoG). After binocular correction and edge detection, the left and right image epipolar lines are horizontally aligned, and the rope edge features in the image are more obvious. Then, feature matching is performed on the left and right images. Feature matching can generally be divided into three steps: the first is feature extraction, that is, extracting key points in the image as feature points of the image; the second is feature description, that is, using vectors to describe the image information of feature points, requiring feature vectors to correspond to feature points one by one; the third is feature matching, that is, calculating the distance between feature vectors. Corner points play a very important role in the analysis and understanding of images. While retaining important features of the image, they can effectively reduce the amount of information data, increase the speed of calculation, and increase the reliability of image matching. Harris operator is a point feature extraction operator based on signals. Inspired by the autocorrelation function in signal processing, a matrix M associated with the autocorrelation function is given. M is a real symmetric matrix, and its eigenvalue is the first-order curvature of the autocorrelation function. If both curvature values ​​are high, the point is considered to be a point feature:

[0164]

[0165] First, a local detection window is set, which is rectangular or circular. When it moves from left to right and from top to bottom on the image, the grayscale value f(x, y) corresponding to the pixel point (x, y) in the detection window area will change. When it exceeds a certain threshold, the central pixel point of the window is recorded as a corner point. The Harris corner point algorithm can be implemented as follows: (1) Calculate the gradient Ix and Iy of the image I(x, y) in the X and Y directions; (2) Calculate the product of the gradients in the two directions of the image; (3) Use the Gaussian function to Gaussian weight Ix2, Iy2 and Ixy (take σ = 1) to generate the elements A, B and C of the matrix M; (4) Calculate the Harris response value R of each pixel, and set R less than a certain threshold t to zero; (5) Perform non-maximum suppression in a 3×3 neighborhood, and the local maximum point is the corner point in the image.

[0166] The present invention uses 3D laser vision equipment HITHMAIC-PSC-VIS-3D to monitor the position of the wire rope. The project uses 4 sets of visual equipment installed at the highest point and the front end of the elephant's trunk, two on each side. The camera and the equipment are connected by angle steel welding. The plane distance of the camera is 40cm to 60cm from the detected wire rope. In order to protect the key equipment of vision and laser light, a protective cover is installed. The protective cover is made of stainless steel to prevent corrosion.

[0167] In order to test the function of the visual system, the present invention completed the testing of four sets of equipment from the perspective of equipment parameters and equipment functions. The test results are as follows:

[0168] Starting from the device itself, the basic parameters of the camera were mapped and corrected. The results are as follows:

[0169] At the same time, the laser wavelength, feature display function, ROI setting function, frame rate device function, image detection function, exposure setting function and gain setting function were tested, and the test results were all qualified.

[0170] The above is only a preferred implementation of a port gantry crane inspection management system. The protection scope of a port gantry crane inspection management system is not limited to the above embodiment. All technical solutions under this idea belong to the protection scope of the present invention. It should be pointed out that for those skilled in the art, several improvements and changes without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A port gantry crane inspection management system, characterized by: The system comprises: Vibration detection system, which performs vibration detection through bearing vibration detection sensor collection equipment; The wire rope flaw detection system can realize the comprehensive detection of wire rope broken wires, wear, corrosion, deformation, loose strands and other defects through the wire rope flaw detection sensor collection equipment, and accurately measure the internal broken wires; Automatic lubrication system, automatic lubrication through automatic lubrication system quantitative distribution valve and controller; The visual positioning system for key parts of the wire rope performs flaw detection through visual system laser and camera equipment.

2. The system according to claim 1, characterized in that: The vibration detection system is directly connected to the corresponding control equipment through the integrated acceleration vibration sensor to achieve online monitoring. With the help of the bus integrated communication system, the vibration signal under motion interference can be effectively analyzed based on high-order signal means. The integrated acceleration vibration sensor integrates the acceleration vibration measuring element and the transmitter module into one, and can directly output the acceleration value signal linearly. The motor equipment status monitoring method based on wavelet analysis and neural network uses wavelet transform to extract the vibration signal characteristics, uses machine learning to identify the characteristics, and outputs the equipment status; uses wavelet time-frequency analysis technology to eliminate noise and filter the motor vibration signal, obtains the frequency band energy through wavelet packet decomposition coefficients, and extracts fault characteristics according to the changes in the energy of each frequency band; Collect vibration sample signals and test signals respectively, use wavelet packets to perform offline analysis and feature extraction on the signals, and obtain sample signal feature vectors and test signal feature vectors respectively; use machine learning or threshold methods to determine the device status.

3. The system according to claim 2, characterized in that: The wire rope flaw detection system uses multiple groups of flaw detection sensor equipment to collect the time domain signals of the wire rope magnetic flux at distributed key points, and uses the magnetic flux analysis method and accumulated data knowledge to identify the state of the wire rope and obtain a comprehensive damage image of the wire rope.

4. The system according to claim 3, characterized in that: The automatic lubrication system completes the interactive control process of lubrication through the oil supply controller PLC and the upper industrial computer via TCP / IP. The lubrication plan is manually formulated in the upper controller. When the lubrication plan is completed and released, the upper controller sends the lubrication time and lubrication amount to the PLC at a regular interval. The PLC receives the instructions to complete the closed-loop control process.

5. The system according to claim 4, characterized in that: The visual positioning system for key parts of wire ropes adopts two detection modules, each of which contains a specially developed optical spot inspection function module for online monitoring of the operation status of wire ropes. The online monitoring optical spot inspection function module can resist extreme sunlight conditions and working vibration interference, and adopts precise optical circuits and accurately developed visual measurement technology to study the changes in the optical properties of the detection object. The array eddy current uses the array eddy current sensor to establish a three-dimensional model for hidden defect detection, optimizes the detection signal space between the center of the detection coil and the detection object, studies the defect signal attenuation trend in the weak magnetic field area between adjacent channels of the array, and adopts an array-placed detection method to realize the crystal oscillator acquisition of detection signal characteristics of defects of different sizes.

6. A method for detecting the state of a wire rope based on vision, characterized in that: The method comprises the following steps: Collect the original wire rope image and pre-process the original image; Perform feature analysis on the preprocessed image to extract the features of the target; According to the extracted target features, conditions are set to perform positioning and settlement. After the original image is processed through a series of algorithms such as preprocessing, feature analysis, spatial positioning, and key spatial response, the spatial position of the wire rope and pulley is obtained. Finally, based on the spatial positioning, the normal and abnormal data of the state are directly obtained through logical reasoning.

7. The method according to claim 6, characterized in that: The image processing process mainly includes two steps: binocular correction and edge detection. The principle of image correction is to reproject the image planes in two different directions to the same plane through matrix transformation, while ensuring that the relative positions of the optical centers of the two cameras remain unchanged, and the optical axes are parallel to each other. In this way, the epipolar lines of the two cameras are horizontally aligned, and matching only requires searching for corresponding points along the horizontal epipolar line direction. When the wire rope features are not obvious, edge detection is used to make the rope edge contour obvious and reduce the feature matching error rate; after binocular correction and edge detection, the left and right image polar lines are horizontally aligned, the rope edge features in the image are more obvious, and then feature matching is performed on the left and right images.

8. The method according to claim 7, characterized in that: feature Matching is divided into 3 steps:

1. Feature extraction, that is, extracting key points in the image as feature points of the image; 2. Feature description, that is, using vectors to describe the image information of feature points, requiring that the feature vectors correspond to the feature points one by one; 3. Feature matching, that is, calculating the distance between feature vectors.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the method according to claims 6 to 8.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the method of claims 6-8 is implemented.

Images