Crane work monitoring method and application thereof in crane work early warning

Abstract

The invention discloses a crane work monitoring method and application thereof in crane work early warning, and relates to the technical field of state monitoring. Comprising the steps that a track data chain of a hoisted object in a hoisting area is obtained according to a hoisting instruction, the track data chain is used for representing motion parameters of the hoisted object in the complete process from hoisting to placing, and feature data content of crane running part abnormity and synchronization losing with an operation instruction is recognized based on the track data chain and hoisting operation dynamic variables. According to the method, the track deviation, the component state deviation, the operation instruction difference and the high-frequency vibration characteristics are uniformly fused, the structure fault, the execution mechanism fault, the instruction asynchronous abnormity and the environment disturbance abnormity are distinguished, and the hidden structure damage caused by long-term use can be identified; in addition, multi-factor complex risks caused by operation impact factors can be distinguished, abnormal execution behaviors caused by inconsistency of operation instructions are marked independently, and therefore the accuracy and pertinence of abnormal recognition are ensured.

Description

Technical Field

[0001] This invention relates to the field of condition monitoring technology, specifically to a crane operation monitoring method and its application in crane operation early warning. Background Technology

[0002] The operational safety of cranes is directly related to the life safety of on-site personnel. In the daily high-frequency lifting, lowering, rotating, and luffing operations, cranes are subjected to repeated heavy loads, impact loads, or harsh environments (such as high temperature, humidity, corrosion, and dust). Key components (such as wire ropes, hooks, hydraulic cylinders, braking systems, and metal structures) are prone to progressive wear, fatigue cracks, loosening, or performance degradation. If these hidden dangers are not detected in time, they may lead to catastrophic accidents such as wire rope breakage, hook detachment, brake failure, and structural collapse, resulting in serious consequences such as casualties, equipment damage, and project shutdown. Therefore, it is necessary to design a crane operation monitoring method.

[0003] A search revealed Chinese invention patent application CN114572846A, which proposes a "wireless transmission method, controller, system, and crane for monitoring data." This method involves sending data acquisition signals to a wireless monitoring device installed on the crane boom via a wireless connection; receiving monitoring data returned by the wireless monitoring device; and obtaining monitoring results based on the monitoring data. Thus, the controller and the wireless monitoring device transmit data wirelessly, achieving stable data transmission, reducing data transmission costs, and improving the safety of crane operations.

[0004] However, in actual use, the aforementioned disclosed devices and similar prior art, when the operating commands and the actual operating state are out of sync during crane operation, will cause the crane's running parts to experience additional frictional forces or stress concentration. However, the aforementioned disclosed methods and similar prior art methods mainly rely on preset command processes for data acquisition, making it difficult to match the dynamic deviations in actual operation in real time. This results in early anomalies such as slight jamming or speed response lag caused by the asynchrony between operation and execution being difficult to effectively identify. This lack of identification ability, in turn, affects the accurate assessment of the health status of the crane's running parts, and to a certain extent restricts the service life of the equipment. Summary of the Invention

[0005] The purpose of this invention is to provide a crane operation monitoring method and its application in crane operation early warning, so as to solve the problems mentioned in the background art.

[0006] Firstly, to achieve the above objectives, the present invention provides the following technical solution: a crane operation monitoring method, comprising: The trajectory data chain of the hoisting object in the hoisting area is obtained according to the hoisting instruction, and the trajectory data chain is used to represent the motion parameters of the hoisting object from hoisting to placing; The feature data content of the abnormality of the running component of the crane and the desynchronization of the operation instruction is identified based on the trajectory data chain and the dynamic variable of the hoisting operation; The initial abnormal state of the crane is identified according to the feature data content of the abnormality of the running component of the crane; The initial abnormal state of the crane is corrected according to the feature data content of the abnormality of the running component of the crane and the desynchronization of the operation instruction; The monitoring data report content is generated according to the corrected abnormal state of the crane.

[0007] As further preferred in the technical solution, the trajectory data chain acquisition method comprises: A monitoring coordinate system is established based on the specification attribute of the crane, the X-axis of the monitoring coordinate system is parallel to the cross beam arm of the crane, and the Y-axis is parallel to the load bearing beam of the crane; According to the hoisting instruction, the real-time coordinate value of the hoisting object in the monitoring coordinate system is obtained, and the time node corresponding to each coordinate value is recorded synchronously to form a trajectory point line bundle in time sequence; The attribute parameters of the breakpoints in the trajectory point line bundle are obtained, including the real-time coordinate value and the time node; The attribute parameters of each breakpoint in the trajectory point line bundle and the corresponding time node are integrated to form a trajectory data chain.

[0008] As further preferred in the technical solution, the feature data content identification method comprises: The trajectory data chain and the dynamic variable of the hoisting operation are associated and integrated according to the time node of each trajectory point in the trajectory data chain to construct a fusion data set indexed by time, containing the real-time coordinate value of the hoisting object, the breakpoint attribute parameter and the corresponding dynamic variable; Based on the fusion data set, the deviation value between the actual trajectory point of the hoisting object corresponding to each time node and the theoretical trajectory point predicted according to the dynamic variable is calculated; According to the trajectory deviation dynamic sequence, multi-dimensional abnormal feature extraction and hierarchical determination are performed.

[0009] As further preferred in the technical solution, the abnormality of the running component of the crane includes the moving speed of the load bearing beam, the rotation angle of the cross beam arm and the tension of the traction rope; The moving speed parameter calculation and evaluation method of the load bearing beam comprises: The displacement signal of the load bearing beam is collected and the real-time moving speed is calculated; A theoretical moving speed model of the load bearing beam is constructed; Calculate the speed deviation and extract the static deviation, dynamic change rate, fluctuation frequency and sliding window stability characteristics; According to the speed deviation characteristics, the abnormal level is divided into slight, moderate and severe.

[0010] As a further preferred embodiment of the present technical solution, the rotation angle calculation and evaluation method of the crane beam arm comprises: Collect the real-time rotation angle of the beam arm and calculate the deviation; Construct a theoretical response model of the rotation angle; Extract the rotation angle static deviation, dynamic change rate, response delay, sliding window standard deviation and high-frequency vibration characteristics; According to the abnormal characteristics, the classification determination includes mild abnormality, moderate abnormality and severe abnormality.

[0011] As a further preferred embodiment of the present technical solution, the tension calculation and evaluation method of the traction rope comprises: Collect the rope tension signal and calculate the real-time tension; Collect the rope tension signal and calculate the real-time tension; Extract the static deviation, dynamic change rate, sliding window stability, multi-branch tension difference and high-frequency vibration characteristics; According to the abnormal characteristics, the classification determination includes mild abnormality, moderate abnormality and severe abnormality.

[0012] As a further preferred embodiment of the present technical solution, the initial crane abnormal state correction method comprises: Construct a global abnormality fusion data structure, including: trajectory deviation, component state deviation, high-frequency vibration characteristics, operation instruction difference characteristics, dynamic variable state and historical abnormality cumulative number; Make an initial judgment and merge the abnormalities according to the component level to form a comprehensive component abnormality matrix; Based on multi-source feature correlation analysis, the root cause is comprehensively determined, including: structural failure, actuator failure, instruction desynchronization abnormality and non-structural disturbance abnormality.

[0013] In the second aspect, in order to perfect the above technical solution, the crane operation monitoring method is also applied to the crane operation early warning, and the crane operation monitoring method is used, and comprises: Trajectory deviation warning: based on the trend prediction of the deviation value, when the predicted deviation exceeds the safety threshold set by human, the sound and light warning is triggered; Bearing beam moving speed abnormality warning: based on the speed deviation trend and fluctuation frequency determination, the moderate or severe jamming risk is determined and the warning is triggered; Beam arm rotation angle abnormality warning: based on the beam arm rotation angle change, sliding window standard deviation and high-frequency vibration characteristics, the warning is triggered; Abnormality early warning of tension of traction rope: early warning is triggered based on tension deviation, multi-branch tension difference and high-frequency vibration characteristics; Synthetic multi-source early warning: trajectory deviation, speed deviation, rotation angle deviation and tension deviation and their trend characteristics are fused, and early warning information is output according to abnormality level and cause analysis.

[0014] Compared with the prior art, the present application has the following advantages: The crane operation monitoring method and its application in crane operation early warning can simultaneously collect hoisting object trajectory data, dynamic variables and real-time operation instructions at the beginning of hoisting operation, and synchronously record them in a unified time sequence, so as to reflect the actual operation state of the crane operation in real time, make the relationship between the trajectory, dynamic variables and operation intention more accurate, and capture the slight operation deviation caused by hidden damage such as guide rail wear, beam support wear, boom crack and internal wire breakage in the early stage. Secondly, in the deviation analysis stage, the present application carries out real-time analysis based on trajectory deviation, speed deviation, angle deviation, tension deviation and operation instruction difference, and identifies early abnormalities such as slight sticking, speed response lag and rotation response delay through difference calculation of operation instruction and execution result and combination of change rate and sliding window stability analysis. Since the above-mentioned abnormalities are often accompanied by additional frictional force or stress concentration phenomenon, they have a significant impact on the health of the running parts. The present method effectively makes up for the technical defect of the prior art that relies on fixed process instructions and cannot identify early execution deviation, thereby improving the accurate evaluation ability of the device health state. Thirdly, in the multi-source data fusion and root cause determination process, the present application uniformly fuses trajectory deviation, component state deviation, operation instruction difference and high-frequency vibration characteristics, realizes the differentiation of structure failure, actuator failure, instruction asynchronous abnormality and environmental disturbance abnormality, can not only identify hidden structural damage caused by long-term use, but also distinguish multi-factor complex risks caused by operation impact factors, and independently label abnormal execution behaviors caused by inconsistent operation instructions, thereby ensuring the accuracy and pertinence of abnormality identification. Finally, in the early warning and reporting stage, the present application provides real-time early warning based on the fusion result, displays instruction deviation trend, trajectory deviation trend and high-frequency vibration characteristics through a visual interface, and timely prompts potential risks that may be caused by execution lag, slight sticking or stress concentration. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 The flowchart of the present application method; Figure 2 The schematic diagram of the monitoring coordinate system of the present application; Figure 3 The identification logic diagram of the feature data of the present application; Figure 4 The flow chart for distinguishing the crane carrying beam moving speed calculation and evaluation method of the present application; Figure 5 The flow chart for distinguishing the crane beam arm rotation angle calculation and evaluation method of the present application; Figure 6 The flow chart for distinguishing the crane traction rope tension calculation and evaluation method of the present application. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0017] Before understanding the technical solutions proposed in the present application, it is necessary to make it clear that the use scenarios of the technical solutions are the daily operation state monitoring of the carrying beam, the beam arm and the traction rope of the crane when the crane is used to lift, carry and rotate materials in the industrial scene, and the early warning scenarios of various potential risks such as blockage, speed deviation, angle deviation and tension abnormality caused by long-term wear, operation impact, inconsistent instructions or environmental disturbance.

[0018] As shown in Figure 1 , the present application provides a technical solution: a crane operation monitoring method, comprising: step S100-step S500.

[0019] Step S100: obtaining the trajectory data chain of the hoisted object in the hoisting area according to the hoisting instruction.

[0020] It is worth noting that the trajectory data chain is used to represent the motion parameters of the hoisted object from the lifting to the placing process.

[0021] As a preferred embodiment, it can be known from Figure 2 that in the present embodiment, the method for obtaining the trajectory data chain comprises: step S101-step S104.

[0022] Step S101: establishing a monitoring coordinate system based on the crane specification attributes, the X axis of the monitoring coordinate system being parallel to the crane beam arm and the Y axis being parallel to the crane carrying beam.

[0023] Step S102: obtaining the real-time coordinate values (X, Y) of the hoisted object in the monitoring coordinate system according to the hoisting instruction, and synchronously recording the time nodes corresponding to each coordinate value to form a time-ordered trajectory point bundle.

[0024] It is worth noting that in the technical solution proposed in the present application, the recording of the time node is realized by the timing unit of the PLC (Programmable Logic Controller) arranged in the crane control cabinet.

[0025] Step S103: Obtain the attribute parameters of the breakpoints in the trajectory point bundle, including real-time coordinate values and time nodes.

[0026] It is worth noting that in the technical solution proposed in the present application, the breakpoints in the trajectory point bundle are used to indicate the deviation of the hoisted object due to the abnormality of the crane running component and the desynchronization of the crane running component and the corresponding operation instruction.

[0027] Step S104: Integrate the attribute parameters of the breakpoints in the trajectory point bundle and the corresponding time nodes, and form a trajectory data chain.

[0028] Step S200: Identify the feature data content of the crane running component abnormality and the desynchronization with the operation instruction based on the trajectory data chain and the dynamic variables of the hoisting operation.

[0029] It should be noted that in the technical solution proposed in the present application, the dynamic variable is used to represent the state parameter directly related to the movement of the hoisted object during the hoisting operation, including the weight of the hoisted object, the lifting speed of the crane, the rotation angle of the beam arm, the moving speed of the load beam, and the tension of the traction rope.

[0030] Specifically, in the technical solution, the method for obtaining the dynamic variable includes: The method for obtaining the weight of the hoisted object is to integrate a weight sensor at the connection part of the crane hook and the traction rope, and the hoisted object weight data obtained by the weight sensor is transmitted to the analog input module of the PLC through the ModbusRTU protocol, to generate a weight dynamic variable sequence indexed by time nodes in real time; The method for obtaining the lifting speed of the crane is to install an absolute rotary encoder at the output shaft end of the crane drive motor, the absolute rotary encoder detects the number of motor revolutions and direction, combines the reduction ratio of the lifting mechanism and the effective diameter of the drum, and calculates the real-time value of the lifting speed, the formula is: lifting speed = (encoder pulse number x drum effective diameter x π) / (encoder resolution x lifting mechanism reduction ratio x 60), where the encoder pulse number is the detection value per unit time, the resolution is the number of pulses per revolution of the encoder, and the calculation result is stored in the speed variable register of the PLC synchronously; The rotation angle acquisition method of the beam arm is that an inclination sensor is installed at the rotation support of the crane beam arm, the inclination sensor detects the angle change between the beam arm and the horizontal plane, outputs an analog voltage signal of 0-5V, corresponding to a rotation angle of 0-360°, and the rotation angle data is processed by the PLC to form a dynamic variable of the rotation angle of the beam arm; The moving speed acquisition method of the bearing beam is that a linear displacement sensor is installed at the sliding connection guide rail of the crane bearing beam and the beam arm, the measuring rod of the linear displacement sensor is fixed to the beam arm, the magnetic ring moves synchronously with the bearing beam, the moving speed of the bearing beam is output in real time by detecting the position change rate of the magnetic ring, the moving speed signal is transmitted to the PLC through the Profibus-DP bus to generate a dynamic variable of the moving speed of the bearing beam; The tension acquisition method of the traction rope is that any one of the tension sensors is arranged at the fixed end (drum side) or guide pulley of the traction rope, the tension sensor detects the pressure of the rope on the pulley, combines the wrap angle of the pulley, calculates the real-time tension of the rope through the formula: tension + (rope pressure on pulley x cos (α / 2)) / (2 x sin (α / 2)), where α is the wrap angle (rad) of the rope on the pulley, and the calculation result is stored as a dynamic variable of the tension of the traction rope by the PLC.

[0031] It is worth noting that the above dynamic variables are all stored by the PLC, and each dynamic variable includes a corresponding time node for synchronization with the track data chain.

[0032] It should be further supplemented that, referring to Figure 3 It can be seen that, in the technical solution proposed in the present application, the feature data content method for identifying the abnormality of the running components of the crane and the desynchronization with the operation instruction includes steps S201-S203.

[0033] Step S201: Based on the time nodes of each track point in the track data chain, the track data chain and the lifting operation dynamic variables are associated and integrated according to the time nodes to construct a fusion data set with time as the index, containing the real-time coordinate values of the hoisted object, the breakpoint attribute parameters and the corresponding dynamic variables.

[0034] Step S202: Based on the fusion data set, the deviation value between the actual track point (X, Y) of the hoisted object corresponding to each time node and the theoretical track point (Xpred, Ypred) predicted according to the dynamic variables is calculated.

[0035] It should be noted that the deviation value includes the lateral deviation ΔX=X-Xpred and the longitudinal deviation ΔY=Y-Ypred, wherein the prediction method of the theoretical track point is: according to the dynamic variables such as the hoisting speed, the rotation angle of the beam arm, the moving speed of the bearing beam, the definition of the monitoring coordinate system, and the kinematic model formula: X pred = X0+ (beam moving speed x Δt) + (Δθ x L x cosθ0); Y pred = Y0+ (lifting speed x Δt) + (Δθ x L x sinθ0); Notably, X0, Y0 are initial coordinates of the hoisted object when hoisting (taken from the starting point of the trajectory data chain), Δt is the cumulative time from the start of hoisting to the current time node (provided by the PLC timing unit), Δθ is the change in the rotation angle of the beam arm (the difference between the current angle and the initial angle), L is the effective length of the beam arm (taken from the crane specification attribute), θ0 is the initial rotation angle of the beam arm (taken from the initial detection value of the inclination sensor), and the calculated lateral deviation ΔX and longitudinal deviation ΔY are stored in the PLC synchronously to form a time-indexed trajectory deviation dynamic sequence.

[0036] Step S203: According to the trajectory deviation dynamic sequence, multi-dimensional abnormal feature extraction and hierarchical judgment are performed.

[0037] Specifically, first, the time domain features of the time series of the lateral deviation ΔX and the longitudinal deviation ΔY are calculated, i.e., the sliding window mean (the window size is 1 / 5 of the typical period of the hoisting operation, such as 5s, used to reflect the continuous trend of the deviation), the absolute maximum (used to identify sudden extreme deviation), the standard deviation (used to measure the dispersion degree of the deviation), and the first-order difference (ΔX=ΔX(t)-ΔX(t-1), ΔY=ΔY(t)-ΔY(t-1), used to represent the change rate of the deviation). Secondly, based on the crane specification attribute (hoisting precision requirement) and the operation safety standard, the hierarchical threshold is set artificially, including the static deviation threshold (ΔX absolute value exceeds ±5cm, ΔY absolute value exceeds ±3cm), the dynamic change rate threshold (ΔX exceeds ±2cm / s, ΔY exceeds ±1.5cm / s), and the statistical stability threshold (such as the sliding window standard deviation exceeds 1.2cm). When any feature exceeds the corresponding threshold, the time node is marked as a potential abnormal point. Then, for the time node corresponding to the potential abnormal point, the dynamic variables in the data set (such as the tension of the traction rope, the difference between the lifting speed and the command speed, and the rotation angle response delay of the beam arm) are associated and fused, and the source of the abnormality is analyzed based on the operation command at the time node, Specifically, if the dynamic variable of the tension of the traction rope exceeds the artificially preset safe range (e.g., the tension is less than 80% of the rated value or higher than 120%), and the deviation of the lifting speed or the moving speed of the carrying beam at the corresponding time node from the operation instruction exceeds the dynamic change rate threshold, it is determined that it is a precursor abnormality of slackening or breaking of the traction rope, which belongs to the abnormality feature of the running component. If the difference between the lifting speed and the instruction speed exceeds ±10%, and the rotation angle of the cross beam arm is delayed in response (i.e., the time difference between the dynamic variable of the rotation angle and the instruction angle exceeds 0.5s), and the first-order difference of the trajectory deviation shows a sudden trend, it is determined that it is an instruction out-of-sync abnormality of the driving motor or the control loop, which belongs to the out-of-sync feature of the operation instruction and the execution component. If the difference between the dynamic variable of the moving speed of the carrying beam and the instruction speed continuously exceeds ±15%, and the output signal fluctuation frequency of the linear displacement sensor is higher than the normal range (more than 5Hz), it is determined that it is a running component abnormality feature of the carrying beam sliding guide jamming or driving mechanism failure.

[0038] Step S300: identifying the initial crane abnormality state according to the feature data content of the crane running component abnormality.

[0039] Specifically, the crane running component abnormality in the embodiment is specifically divided into: the moving speed of the crane carrying beam, the rotation angle of the cross beam arm, and the tension of the traction rope.

[0040] As a supplement to the content of step S300, refer to Figure 4 It can be known that, in the embodiment, the moving speed parameter calculation and evaluation method of the crane carrying beam includes: steps S301-S304 Step S301: collecting the carrying beam displacement signal and calculating the real-time moving speed.

[0041] In the embodiment, the carrying beam moving speed is obtained based on the linear displacement sensor in step S200. The PLC reads the displacement value S(t) of the linear displacement sensor at a fixed sampling period (10ms), and calculates the moving speed Vhob(t) of the carrying beam according to the discrete difference formula, the calculation formula is: Vhob(t)=[S(t)-S(t-1)] / Δt, wherein Δt is the sampling period of the PLC, and the calculated real-time value of the carrying beam moving speed is written into the speed variable register of the PLC synchronously.

[0042] Step S302: constructing a theoretical response model of the moving speed of the carrying beam.

[0043] It is worth noting that in step S302, by means of the target moving speed instruction Vcmd(t) of the carrying beam in the operation instruction, combined with the crane specification attributes (including the guide rail friction coefficient μ, the carrying beam rated load Massr, and the drive mechanism rated output force Fdr), a theoretical speed model of the carrying beam is constructed in the PLC for deriving the theoretical moving speed Vpred(t) of the carrying beam.

[0044] It is worth noting that in step S302, the theoretical speed model formula is: Vpred(t)=Vcmd(t)-f(μ,Massr,Fdr)×k, wherein f(μ,Massr,Fdr) is the speed loss amount caused by factors such as friction and structural load, and k is a structural compensation coefficient (obtained based on experimental calibration), and the PLC stores the calculated Vpred(t) as a theoretical speed dynamic variable for subsequent deviation analysis.

[0045] Step S303: Calculate the moving speed deviation of the carrying beam and extract the speed abnormality feature.

[0046] It should be noted that in step S303, the speed deviation ΔVhob(t) is calculated based on the real-time speed Vhob(t) and the theoretical speed Vpred(t), and the calculation formula is: ΔVhob(t)=Vhob(t)-Vpred(t).

[0047] As a supplement to the calculation formula of step S303, in order to identify the abnormal motion of the carrying beam, the following features are calculated for the speed deviation sequence: First, the static deviation feature: the absolute value of ΔVhob(t) exceeds the preset threshold (such as ±0.15 m / s); Second, the dynamic change rate feature: the first-order difference ΔVhob'(t)=ΔVhob(t)-ΔVhob(t-1) exceeds the change rate threshold (such as ±0.05 m / s²); Third, the fluctuation frequency feature: based on the short-time Fourier transform (STFT) in the prior art, the local frequency component of the carrying beam speed deviation is calculated, and when the frequency component exceeds 5 Hz, it is determined as a guide rail vibration abnormality; Fourth, the sliding window stability feature: the standard deviation σ is calculated with a window of 3-5s, and when σ is higher than the preset stability threshold (such as 0.08 m / s), it is marked as a stability abnormality.

[0048] Step S304: Determine the abnormal state of the carrying beam based on the speed deviation feature.

[0049] It should be noted that in step 304, based on the synchronous association of the speed data with the operation instruction and the dynamic variable, according to the abnormal feature triggering condition, the following hierarchical determination is performed: First, slight jamming (first-level abnormality): ΔVhob(t) continuously exceeds the static deviation threshold but does not exceed the rate of change threshold; the fluctuation frequency is between 2-3Hz, at which point it is determined that the bearing beam guide rail has slight wear or local damping increase.

[0050] Second, moderate jamming or insufficient drive compensation (Level 2 anomaly): When the absolute value of ΔVhob(t) exceeds the static deviation threshold and a sudden change in the rate of change occurs, and the fluctuation frequency is between 3-5Hz, the phenomenon of excessive current in the drive motor (provided by the dynamic variable of the actuator) also occurs. At this time, it is determined to be a moderate jam caused by slider wear, insufficient lubrication, local track contamination, etc.

[0051] Severe jamming or drive mechanism failure (Level 3 anomaly): ΔVhob(t) exceeds the dynamic rate of change threshold multiple times and exhibits a sudden pattern. At this time, the frequency component obtained by analysis is significantly higher than 5Hz. The operating command speed Vcmd(t) is normal, but the speed of the load-bearing beam has almost no response (Vhob is close to 0). At this time, it is judged as a major anomaly, including severe deformation of the sliding guide rail, misalignment of the drive mechanism gears, and broken transmission chain, which require immediate shutdown for handling.

[0052] As a preferred implementation scheme, refer to Figure 5 The method for calculating and evaluating the rotation angle of the crane boom in step S300 includes steps S311-S315.

[0053] Step S311: Collect the real-time rotation angle of the crossbeam arm and calculate the deviation Δθ.

[0054] Specifically, in this implementation scheme, the rotation angle of the beam arm is provided by an inclination sensor installed at the rotating support. The PLC reads the voltage signal U(t) of the inclination sensor at a fixed sampling period and calculates the rotation angle θ(t) according to the linear mapping formula. The calculation formula is: θ(t) = U(t) × (360° / 5V). The real-time value of the rotation angle is synchronously written into the PLC's angle dynamic variable register for subsequent deviation analysis and trajectory prediction calculation.

[0055] Step S312: Construct a theoretical response model for the rotation angle.

[0056] According to the beam arm target angle instruction θcmd(t) in the operation instruction, a beam arm theoretical rotation response model is established in combination with beam arm structure parameters (including beam arm effective length L, rotation damping coefficient Cd, and driving torque Md), and a beam arm theoretical rotation angle value θpred(t) is derived, wherein the theoretical rotation angle model can be expressed as: θpred(t)=θcmd(t)-g(Cd,Md,L)×β, wherein g(Cd,Md,L) is an angle loss caused by rotation damping, mechanism inertia and load change, and β is a structure compensation coefficient obtained through experimental calibration. The PLC stores the calculated θpred(t) as a theoretical rotation angle dynamic variable.

[0057] It should be noted that the beam arm effective length L is an equivalent distance between the rotation center of the beam arm and the action point of the hoisted object, and the value thereof is provided by crane product specification parameters. Specifically, the beam arm effective length L is obtained by querying the crane manufacturer's technical manual, the equipment nameplate or the equipment factory configuration file, and the PLC reads and stores the fixed parameter in the initialization stage.

[0058] The rotation damping coefficient Cd is mainly used to represent the damping amount of the beam arm in the rotation motion due to factors such as friction, structural resistance and lubrication state, and the value thereof is difficult to be directly measured by hardware, and is thus obtained through experimental calibration.

[0059] The calibration steps include: driving the beam arm to rotate at a fixed angular velocity in an empty state, recording the change curve of the angle with time through the inclination sensor, comparing the response with the theoretical undamped model, calculating the angular velocity attenuation rate, and substituting the attenuation rate into the damping dynamics model to inversely solve the rotation damping coefficient Cd.

[0060] The driving torque Md is obtained in the following manner: the driving torque Md is determined by the rated output characteristic of the beam arm driving motor, and is clear in the equipment manufacturing stage, and the value thereof is obtained in the following manner: for a fixed output level driving mechanism, Md is derived from the motor model technical manual or the equipment factory technical parameter, and is loaded by the PLC in the initialization stage; for a variable torque mechanism affected by load change, the PLC calculates the real-time driving torque Md(t)=Kt×I(t) according to the real-time current value I(t) of the driving motor and the motor torque constant Kt; for the beam arm theoretical response model in the embodiment, the rated driving torque Md or the average value of Md(t) calculated in real time is used by default to ensure the stability of the model.

[0061] Step S313: Calculate the rotation angle deviation and perform feature extraction.

[0062] Based on the real-time rotation angle θ(t) and the theoretical rotation angle θpred(t), the rotation angle deviation Δθ(t) is calculated, and the calculation formula is: Δθ(t)=θ(t)-θpred(t). On this basis, the rotation angle deviation sequence is extracted as follows: Static deviation feature: |Δθ(t)| exceeds the preset static deviation threshold (such as ±2°).

[0063] Dynamic change rate feature (used to identify sudden or jittering abnormalities): Δθ'(t)=Δθ(t)-Δθ(t-1), when the absolute value of Δθ'(t) exceeds the change rate threshold (such as ±0.5° / s), it is marked as a potential abnormality.

[0064] Response delay feature (used to identify instruction desynchronization): if θ(t) is delayed compared to θcmd(t) and the delay time exceeds 0.5s, it is determined as a response lag feature.

[0065] Stability feature: based on a sliding window (such as 3-5s), the standard deviation σθ is calculated, and when σθ exceeds the stability threshold (such as 1.0°), it is determined as a rotating arm jittering abnormality.

[0066] High-frequency fluctuation feature (used to detect rotating bearing or support wear): based on the short-time Fourier transform STFT in the prior art to analyze the deviation signal, if the frequency component exceeds 4Hz, it is determined as a mechanism high-frequency vibration abnormality.

[0067] Step S314: Abnormal pattern recognition combined with dynamic variables and operation instructions.

[0068] According to the dynamic variables (including lifting speed, load beam moving speed, traction rope tension, etc.) and operation instructions obtained in step S200, the abnormal reasons are analyzed and classified: Insufficient driving torque or reducer wear causes insufficient rotation response: persistent static deviation Δθ(t)>±2°, dynamic variable driving motor current continuously high, rotation angle response delay>0.5s.

[0069] Rotating support or bearing wear causes jittering abnormality: σθ (sliding window standard deviation) exceeds the threshold; STFT frequency component>4Hz; Δθ'(t) exists periodic fluctuation.

[0070] Operation instruction and execution component desynchronization: θcmd(t) and θ(t) time difference exceeds 1.0s, Δθ'(t) mutation accompanied by trajectory deviation ΔX, ΔY first-order difference abnormality, cross beam arm driving mechanism feedback signal lag or fluctuation.

[0071] Load change causes rotation deviation abnormality (such as hoisting object swing); The tension of the traction rope deviates from the rated range, the rotation angle deviation occurs simultaneously with the rapid change of the lifting speed, and the deviation characteristics are non-structural fluctuations (random fluctuations).

[0072] Step S315: hierarchical determination and alarm output based on abnormal characteristics.

[0073] To realize rapid diagnosis of the rotating mechanism, the present embodiment gives three types of hierarchical rules: First level (mild abnormality): Δθ(t) is slightly over-standard but has no high-frequency component, which belongs to normal wear or short-term interference.

[0074] Second level (moderate abnormality): σθ is high, Δθ'(t) fluctuates obviously, and there is a response delay, which requires maintenance.

[0075] Third level (serious abnormality): θ(t) and θcmd(t) are seriously out of synchronization, high-frequency vibration is significant, the rotation angle has almost no response or appears sudden deviation, at this time, immediate shutdown is required to prevent the rotation of the beam arm from getting out of control and causing lifting accidents.

[0076] As a preferred embodiment, reference is made to Figure 6 For the tension calculation and evaluation method of the traction rope in step S300, steps S321-S325 are included.

[0077] Step S321: collect the rope tension signal and calculate the real-time tension.

[0078] In the present embodiment, the tension of the traction rope is provided in real time by a tension sensor installed at the fixed end or guide pulley of the rope, and the PLC reads the pressure signal P(t) of the tension sensor at a fixed sampling period, and calculates the rope tension T(t) according to the pulley wrap angle α, the calculation formula is: T(t)=(P(t)×cos(α / 2)) / (2×sin(α / 2)), wherein P(t) is the pressure of the rope acting on the pulley, α is the rope wrap angle (rad), which is provided by the crane structure parameters, T(t) is the real-time tension of the traction rope, which is written into the PLC as a dynamic variable of tension, it is worth noting that if the crane has multiple rope branches, this scheme supports calculating the tension of each branch respectively, and storing it in the form of a linked list in the PLC, which is convenient for subsequent abnormality determination.

[0079] Step S322: construct a theoretical tension model of the traction rope.

[0080] It should be noted that in step S322, a theoretical tension model is constructed according to the weight of the hoisted object W(t), the lifting speed Vh(t), the crane structure damping coefficient ζ, and the rope self-weight Wr parameters, which is used to derive the theoretical tension Tpred(t) at the corresponding time node.

[0081] Notably, the theoretical model is represented as: Tpred(t) = W(t) / n + Wr + h(Vh(t), ζ), where: n is the number of rope branches, Wr is the tension component generated by the self-weight of the rope, h(Vh(t), ζ) is the additional load caused by the hoisting acceleration, speed fluctuation and structural damping, which is obtained by experimental calibration, and PLC stores the calculated Tpred(t) as a theoretical tension dynamic variable for deviation analysis.

[0082] Step S323: Calculate the tension deviation and extract the abnormal features.

[0083] Based on the real-time tension T(t) and the theoretical tension Tpred(t), the tension deviation AT(t) is calculated: AT(t) = T(t) - Tpred(t).

[0084] In order to identify the rope anomaly, the following features are extracted from the deviation sequence: Static tension deviation feature (under-tension or over-tension): |AT(t)| > preset static threshold (such as ±20% of the rated tension).

[0085] Dynamic tension change rate feature (used to detect instantaneous impact, loose rope or slipping): AT'(t) = AT(t) - AT(t-1), when AT'(t) exceeds the change rate threshold (such as ±15% of the rated tension / s), it is marked as a dynamic anomaly.

[0086] Sliding window stability feature (used to detect jitter): calculate the standard deviation σT for a 3-5s window, when σT > stability threshold (such as 10% of the rated tension), it is determined as a rope fluctuation anomaly.

[0087] Frequency domain feature (used to identify rope fatigue or pulley wobble): perform short-time Fourier transform (STFT) on AT(t), when the high-frequency component > 6Hz and the amplitude exceeds the set threshold, it is considered that there may be rope fatigue, pulley bearing damage or structural problems.

[0088] Multi-branch tension difference feature (used to identify uneven stress on the rope): calculate the difference for multiple rope branch tensions Ti(t): ΔTimb(t) = Ti(t) - Tj(t), if |ΔTimb(t)| exceeds the balance threshold (such as 10% of the rated tension), it is determined that the rope winding is uneven or locally stressed.

[0089] Step S324: Abnormal reason analysis and classification combined with dynamic variables.

[0090] It should be noted that in step S324, the tension deviation features are associated with dynamic variables such as hoisting speed, rotation angle, and moving speed of the load beam to determine the root cause of the anomaly: Rope slack or slip: T(t) is significantly lower than 80% of the rated tension, ΔT'(t) appears rapid decline, and the absolute value of the hoisting speed Vh(t) decreases or produces jitter, followed by a sudden change in the trajectory deviation ΔY.

[0091] Load mutation or impact load: ΔT'(t) suddenly increases and is accompanied by high-frequency vibration, the hoisting speed suddenly changes (caused by load swing), and the multi-branch tension difference ΔTimb(t) is abnormal.

[0092] Pulley block or guide mechanism wear: σT jitter is significantly more than 10%, and the high-frequency component (>6Hz) of STFT is significant.

[0093] Overload risk: T(t) continuously exceeds 120% of the rated value, Tpred(t) is significantly higher than W(t) calculated based on the weight sensor, the hoisting speed decreases, and the motor current is high (from the dynamic variable).

[0094] Step S325: Hierarchical determination and safety alarm based on abnormal characteristics.

[0095] In order to realize real-time monitoring of the rope state, the following three-level determination rules are set in the embodiment: First level (mild abnormality): ΔT(t) is slightly exceeded but has no continuous trend, and σT is high but does not trigger the dynamic threshold value, at this time, the rope is slightly loose or disturbed for a short time, and can be continuously observed.

[0096] Second level (moderate abnormality): ΔT(t) exceeds the static threshold value, ΔT'(t) appears mutation, and the multi-branch tension is uneven, at this time, maintenance needs to be arranged, such as adjusting the rope winding or checking the pulley assembly.

[0097] Third level (serious abnormality): T(t) < 70% of the rated tension (slack risk) or > 130% (serious overload), high-frequency vibration characteristics are obvious (>6Hz), and ΔT'(t) has a continuous abnormal trend, at this time, immediate shutdown is required to prevent rope rupture or hoisting stress out of control.

[0098] Step S400: According to the characteristic data content of the crane running component abnormality and the operation instruction desynchronization, the initial crane abnormal state is corrected.

[0099] Specifically, in the embodiment, step S400 is used to build a comprehensive abnormality determination model to finally identify the overall running state of the crane, specifically including steps S401-S403.

[0100] Step S401: A global abnormality fusion data structure is constructed.

[0101] It should be noted that in step S401, the three component abnormalities (carrying beam moving speed abnormality, cross beam arm rotation angle abnormality, traction rope tension abnormality) identified in step S300 are fused with the trajectory deviation abnormality features (ΔX, ΔY, ΔX', ΔY') extracted in steps S200-S203 and the operation instruction desynchronization features to construct a global abnormality fusion data structure Dfuse(t), wherein the global abnormality fusion data structure Dfuse(t) includes: trajectory deviation features at the current time node, current component state deviation features (speed deviation, angle deviation, tension deviation), high-frequency vibration features (frequency components from STFT), operation instruction difference features (such as Vcmd-Vhob, θcmd-θ), dynamic variable state (tension, speed, motor current), historical abnormality cumulative number and duration.

[0102] Step S402: Abnormality preliminary judgment and merging according to component level.

[0103] It should be noted that in step S402, according to the three-level judgment results obtained in step S300, component-level abnormality labels are formed, including: carrying beam moving speed abnormality level: Lhob∈{normal, first level, second level, third level}; cross beam arm rotation angle abnormality level: Lrot∈{normal, first level, second level, third level}; traction rope tension abnormality level: Lrope∈{normal, first level, second level, third level}, the component-level abnormality label time node combines the abnormal states of the three components into a comprehensive component abnormality matrix: Aparts(t)={Lhob(t), Lrot(t), Lrope(t)} and synchronously includes the trajectory abnormality level Ltraj(t), wherein the trajectory abnormality level is derived from the deviation feature threshold judgment in S203.

[0104] Step S403: Root cause comprehensive judgment based on multi-source feature correlation analysis.

[0105] In step S403, the correlation between Apᴀʀᴛs(t), Ltraj(t) and Dfuse(t) is used for root cause inference, and logical judgment is made according to the coupling features between abnormalities, including: (1) Structural root cause judgment: any of the three abnormalities of the carrying beam, cross beam arm or rope triggers, the trajectory deviation continuously abnormally and ΔX', ΔY' exist mutation, and the high-frequency vibration feature (>5-6Hz) continuously exists, at this time it is judged as a structural failure (such as guide rail deformation, bearing damage, rope fatigue).

[0106] (2) Actuator root cause judgment: one of the following conditions occurs: the drive motor current is continuously high , the theoretical speed / angle response is normal but the actual response is insufficient, and the hysteresis ( > 0.5-1.0s) of θcmd-θ or Vcmd-Vhob occurs synchronously, at which time it is determined that the actuator fails (motor torque is insufficient, reducer is abnormal, transmission chain slips).

[0107] (3) Instruction out-of-sync root cause determination: the deviation between the instruction value and the actual value is > 10-15% for a long time , Δθ'(t) or ΔVhob'(t) has a transient mutation, the dynamic variable is normal but the trajectory deviation is significant, at which time it is determined that the operation instruction and the execution component are out of sync.

[0108] (4) Load disturbance or non-structural fluctuation root cause determination: there is: high tension fluctuation σT but no static deviation, random fluctuation of beam arm angle deviation, and swing of lifting speed without structural characteristics, at which time it is determined that the non-structural anomaly is caused by the swing of the hoisted object or transient disturbance.

[0109] Step S500: generating monitoring data report content according to the corrected crane abnormal state.

[0110] Specifically, in step S500, the generation of the monitoring report content is achieved by reading the fused abnormal state Dfuse(t), dynamic variable, trajectory deviation, and operation instruction deviation from the PLC, integrating the data according to the time node, generating a unified table and sequence, and then combining the S400 corrected abnormal state to label the abnormal level and root cause for each time node.

[0111] As a supplement to the above crane operation monitoring method, the technical scheme of the crane operation monitoring method in the application is also applied to the work warning.

[0112] Embodiment one, trajectory deviation warning Before the lifting operation starts, first of all, according to the lifting instruction, the trajectory data chain of the hoisted object in the monitoring coordinate system is obtained, including the X, Y coordinate values corresponding to each time node, and the time node is recorded through the timing unit of the PLC, then the theoretical trajectory coordinates Xpred, Ypred of the hoisted object are calculated according to the dynamic variables of the lifting operation (lifting speed, beam arm rotation angle, carrying beam moving speed, traction rope tension, etc.), and a real-time trajectory deviation sequence ΔX=X−Xpred, ΔY=Y−Ypred is formed, at each time node, the trend of the deviation is analyzed in a sliding window manner, the mean, slope and acceleration are calculated, which are used to identify whether the trajectory deviation is in a continuous upward trend, when the predicted trajectory deviation is in the future 5 to 10 seconds and exceeds the safety threshold (such as ΔX>±5cm, ΔY>±3cm), the early warning is triggered, it is worth noting that the early warning information displays the real-time trajectory deviation curve and the early warning level on the monitor of the crane, at the same time, the PLC sends out sound and light alarm to prompt the operator to take measures such as reducing speed or adjusting the lifting route to avoid the hoisted object from colliding or swinging out of limits.

[0113] Embodiment two, abnormal carrying beam moving speed early warning In the lifting operation, the real-time displacement S(t) of the carrying beam is obtained by installing a linear displacement sensor, and the PLC calculates the moving speed Vhob(t) of the carrying beam with a fixed sampling period Vhob(t)=[S(t)−S(t−1)] / Δt, then, according to the target speed Vcmd(t) of the carrying beam in the operation instruction and the crane specification attributes (guide rail friction coefficient, rated load, drive mechanism output force), a theoretical speed model Vpred(t)=Vcmd(t)−f(μ,Massr,Fdr)×k is constructed, and the speed deviation ΔVhob(t)=Vhob(t)−Vpred(t) is calculated, and the ΔVhob(t) sequence is feature extracted, including static deviation, dynamic change rate, fluctuation frequency and sliding window standard deviation, it is worth noting that if the deviation trend continues to increase and the fluctuation frequency is close to 3 to 5Hz, it is determined that the carrying beam has a moderate jamming risk, triggering a secondary early warning, at this time, the early warning information of the secondary early warning is output through the PLC sound and light alarm, and the speed deviation trend curve is displayed on the monitor of the crane, prompting the operator to reduce the moving speed of the carrying beam or suspend the operation.

[0114] Embodiment three, abnormal beam arm rotation angle early warning In the lifting operation, the rotation angle of the beam arm θ(t) is collected in real time by the inclination sensor, the PLC reads the voltage signal at a fixed sampling period and converts it to an angle value, combines the operation command θcmd(t) and the beam arm structure parameters (effective length, rotation damping coefficient, driving torque), constructs the theoretical rotation angle model θpred(t) = θcmd−g(Cd,Md,L)×β, then calculates the deviation Δθ(t) = θ(t)−θpred(t), subsequently, extracts the static deviation, the rate of change, the sliding window standard deviation and the high frequency vibration features, when the sliding window standard deviation σθ is higher than the threshold value, and the high frequency component > 4Hz, it is judged that the beam arm has a rotation response lag or a jitter abnormal risk, triggering a secondary warning, the warning information of the secondary warning is sent out by the PLC through sound and light alarm, at the same time, the rotation angle deviation trend curve and the high frequency vibration amplitude are displayed on the monitor of the crane, prompting the operator to slow down or pause the rotation operation, and checking whether the rotation support or bearing needs to be maintained.

[0115] Embodiment four, abnormal warning of tension of traction rope In the lifting operation, the tension of the traction rope T(t) = (P(t) × cos(α / 2)) / (2 × sin(α / 2)) is calculated by the PLC, then combined with the weight of the hoisted object W(t), the hoisting speed Vh(t), the rope self weight Wr and the damping coefficient ζ to construct the theoretical tension model Tpred(t) = W / n + Wr+h(Vh, ζ), to calculate the tension deviation ΔT(t) = T(t)−Tpred(t), subsequently, the static deviation, the dynamic rate of change, the sliding window standard deviation, the multi-branch tension difference and the high frequency vibration features are extracted from the sequence ΔT(t), when the deviation trend shows that the tension continues to decrease by more than 20%, or the multi-branch tension difference is abnormal, and the high frequency vibration > 6Hz, any one of secondary or tertiary warning is triggered, at this time, it is prompted that the rope may be slack or there is an overload risk, the PLC sends out sound and light alarm, the monitor of the crane displays the real-time tension and the multi-branch difference curve, the operation suggestions include immediately reducing the load, stopping or checking the rope and the sheave assembly.

[0116] Embodiment five, comprehensive multi-source warning In the hoisting operation, the track deviation (ΔX, ΔY), the speed deviation ΔVhob of the carrying beam, the arm rotation angle deviation Δθ of the cross beam, the rope tension deviation ΔT and its trend characteristics, the high-frequency vibration component, the operation instruction differential characteristics and the dynamic variable state are fused to form a global abnormal fusion data structure Dfuse(t), then, the component level abnormal grades (the carrying beam Lhob, the cross beam arm Lrot, the rope Lrope) and the track abnormal level Ltraj are merged according to the time node, and the root cause correlation analysis is carried out, when the trend of any component or track deviation shows potential safety risk, the early warning level (first level, second level, third level) is determined according to the comprehensive characteristics, and the specific early warning information is output combined with the root cause analysis, the early warning information is visually displayed through the PLC sound and light alarm and the monitor of the crane.

[0117] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications, changes, omissions, substitutions and adaptations can be made by those skilled in the art without departing from the application, which is defined by the following claims.

Claims

1. Crane operation monitoring method, characterized in that, The method comprises the following steps: According to the lifting instruction, the trajectory data chain of the hoisting object in the hoisting area is obtained, and the trajectory data chain is used to represent the motion parameters of the hoisting object from lifting to placing; Based on the trajectory data chain and the dynamic variable of the hoisting operation, the feature data content of the abnormal operation of the crane running component and the out-of-sync operation instruction is identified; According to the feature data content of the abnormal operation of the crane running component, the initial abnormal state of the crane is identified; According to the feature data content of the abnormal operation of the crane running component and the out-of-sync operation instruction, the initial abnormal state of the crane is corrected; According to the corrected abnormal state of the crane, the monitoring data report content is generated.

2. Crane operation monitoring method according to claim 1, characterized in that The method for obtaining the trajectory data chain comprises the following steps: Based on the specification attribute of the crane, a monitoring coordinate system is established, and the X-axis of the monitoring coordinate system is parallel to the cross beam arm of the crane, and the Y-axis is parallel to the load bearing beam of the crane; According to the lifting instruction, the real-time coordinate value of the hoisting object in the monitoring coordinate system is obtained, and the time node corresponding to each coordinate value is recorded synchronously to form a time-ordered trajectory point bundle; The attribute parameters of the breakpoints in the trajectory point bundle are obtained, and the attribute parameters include the real-time coordinate value and the time node; The attribute parameters and the corresponding time nodes of each breakpoint in the trajectory point bundle are integrated to form the trajectory data chain.

3. Crane operation monitoring method according to claim 1, characterized in that: The identification method of the feature data content comprises the following steps: Based on the time node of each trajectory point in the trajectory data chain, the trajectory data chain and the dynamic variable of the hoisting operation are associated and integrated according to the time node to construct a fusion data set indexed by time, containing the real-time coordinate value of the hoisting object, the breakpoint attribute parameter and the corresponding dynamic variable; Based on the fusion data set, the deviation value between the actual trajectory point of the hoisting object corresponding to each time node and the theoretical trajectory point predicted according to the dynamic variable is calculated; According to the trajectory deviation dynamic sequence, multi-dimensional abnormal feature extraction and hierarchical judgment are performed.

4. The crane work monitoring method according to claim 1, characterized by: The abnormal operation of the crane running component includes the moving speed of the crane load bearing beam, the rotation angle of the cross beam arm and the tension of the traction rope; The calculation and evaluation method of the moving speed parameter of the crane load bearing beam comprises the following steps: Collect the displacement signal of the load bearing beam and calculate the real-time moving speed; Construct a theoretical moving speed model of the load bearing beam; Calculate the speed deviation and extract the static deviation, dynamic change rate, fluctuation frequency and sliding window stability features; According to the speed deviation features, hierarchical judgment is performed, and the abnormal level is divided into slight, moderate and severe.

5. Crane operation monitoring method according to claim 4, characterized in that The calculation and evaluation method of the rotation angle of the cross beam arm of the crane comprises the following steps: Collect the real-time rotation angle of the cross beam arm and calculate the deviation; Construct a theoretical response model of the rotation angle; Extract the rotation angle static deviation, dynamic change rate, response delay, sliding window standard deviation and high-frequency vibration features; According to the abnormal features, hierarchical judgment is performed, including slight abnormality, moderate abnormality and severe abnormality.

6. The crane work monitoring method according to claim 4, characterized by: The calculation and evaluation method of the tension of the traction rope comprises the following steps: Collect the rope tension signal and calculate the real-time tension; Collect the rope tension signal and calculate the real-time tension; Extract the static deviation, dynamic change rate, sliding window stability, multi-branch tension difference and high-frequency vibration features; According to the abnormal features, hierarchical judgment is performed and safety warning is output.

7. The crane work monitoring method according to claim 1, characterized by: The correction method of the initial abnormal state of the crane comprises the following steps: Constructing global abnormal fusion data structure, including: trajectory deviation, component state deviation, high-frequency vibration features, operation instruction difference features, dynamic variable state and historical abnormal cumulative number; Abnormal preliminary judgment and merging according to component level to form comprehensive component abnormal matrix; Based on multi-source feature correlation analysis for root cause comprehensive judgment, including: structure fault, actuator failure, instruction desynchronization abnormality and non-structural disturbance abnormality.

8. Use of a crane operation monitoring method in crane operation warning, using any of the crane operation monitoring methods according to claims 1 - 7, characterized in that, Including: Trajectory deviation early warning: based on the trend prediction of deviation value, triggering sound and light early warning when the predicted deviation exceeds the safety threshold set by human; Bearing beam moving speed abnormality early warning: based on the speed deviation trend and fluctuation frequency to determine the risk of moderate or serious jamming and trigger early warning; Crossbeam arm rotation angle abnormality early warning: based on the crossbeam arm rotation angle change, sliding window standard deviation and high-frequency vibration features to trigger early warning; Traction rope tension abnormality early warning: based on the tension deviation, multi-branch tension difference and high-frequency vibration features to trigger early warning; Comprehensive multi-source early warning: fusion of trajectory deviation, speed deviation, rotation angle deviation and tension deviation and their trend features, outputting early warning information according to abnormal level and cause analysis.

Images