A method for early warning of operational risks of cutting machines for cutting cable protection pipes.
By comprehensively collecting and analyzing multi-dimensional sensor signals from the cutting machine, the problem of insufficient monitoring dimensions in existing technologies has been solved, enabling accurate early warning of the cutting machine's operating status and refined judgment of risk situations, thereby improving the operational safety of the cutting machine.
Patent Information
- Application Number
- CN202611078338.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-20
- Publication Date
- 2026-11-17
AI Technical Summary
Existing risk monitoring methods for cutting machines lack the correlation between pipe deformation and feeding conditions, resulting in insufficient monitoring dimensions. This makes it difficult to accurately quantify the overall deviation of equipment operation, leading to missed or false alarms and failing to meet the safety management requirements of cable protection pipe cutting production.
By synchronously collecting drive motor current signals, feed speed signals, pipe wall deformation feedback signals, and spindle vibration signals, component analysis, operating condition calibration, and trend extrapolation are performed to form a quantitative judgment of comprehensive risk deviation. Combined with the quantitative formula, a risk situation stratification judgment is achieved.
It improves the accuracy and efficiency of risk warning for cutting machine operation, enabling precise differentiation of potential equipment operation hazards and reducing the probability of pipe damage and equipment overload failure.
Smart Images

Figure FT_1 
Figure QLYQS_1
Abstract
Description
Technical Field
[0001] This invention relates to the field of equipment monitoring technology, and in particular to a method for early warning of operational risks of a cutting machine for cutting cable protection pipes. Background Technology
[0002] Cable protection pipe cutting operations rely on cutting machines to cut pipes to a fixed length. During the cutting process, abnormal operating conditions such as pipe extrusion deformation, abnormal spindle vibration, feed rate instability, and sudden changes in motor load can easily lead to malfunctions such as cut edge chipping, pipe scrapping, spindle wear, and even equipment overload shutdown. Currently, most mainstream cutting machine risk monitoring methods use single sensor parameters to judge the operating status, relying solely on motor current or vibration data for threshold alarms. This fails to synchronously correlate the coupled effects of pipe deformation and feed conditions, resulting in a significant lack of monitoring dimensions and making it difficult to fully reconstruct the true operating status of the entire cutting process.
[0003] The existing early warning mechanism uses fixed thresholds to determine risk levels, failing to dynamically calibrate the equipment's baseline load based on real-time feed speed and spindle vibration. Changes in cutting conditions lead to significant deviations in the original judgment criteria, resulting in a pronounced lag in load deviation identification. Furthermore, the current solution lacks a collaborative quantitative analysis logic for pipe deformation and motor load, making it impossible to accurately quantify the overall deviation of equipment operation. The fixed and rigid risk range divisions are prone to missed and false alarms, resulting in low overall early warning accuracy and response efficiency. This fails to meet the safety management requirements for continuous and stable cutting production of cable protection pipes. Therefore, improving the efficiency of risk early warning for cutting machines has become an urgent problem to be solved. Summary of the Invention
[0004] This invention provides a method for early warning of operational risks of a cutting machine for cutting cable protection pipes, in order to solve the problems mentioned in the background art.
[0005] To achieve the above objectives, the present invention provides a method for early warning of operational risks of a cutting machine for cutting cable protection pipes, comprising:
[0006] P1. Acquire the drive motor current signal, feed speed signal, pipe wall deformation feedback signal and spindle vibration signal of the target cutting machine, and perform component analysis on the drive motor current signal to obtain the reference load of the target cutting machine.
[0007] P2. Based on the feed speed signal and the spindle vibration signal, the reference load is calibrated according to the working condition to obtain the corrected load of the target cutting machine;
[0008] P3. Sensing the difference trend between the corrected load and the reference load to obtain the load deviation trend of the target cutting machine;
[0009] P4. Based on the pipe wall deformation feedback signal, perform deformation load collaborative analysis on the load deviation trend to obtain the comprehensive risk deviation of the target cutting machine;
[0010] P5. Based on the load deviation trend, perform deviation inversion on the benchmark load to obtain the risk judgment range of the target cutting machine;
[0011] P6. Based on the risk assessment interval, perform risk situation prediction on the comprehensive risk deviation to obtain the risk level of the target cutting machine.
[0012] In a preferred embodiment, acquiring the drive motor current signal, feed speed signal, pipe wall deformation feedback signal, and spindle vibration signal of the target cutting machine includes:
[0013] Collect raw sensor data from the target cutting machine;
[0014] The original sensor data is filtered to remove interference, resulting in the denoised sampling data of the target cutting machine.
[0015] The denoised sampling data is sorted to obtain the drive motor current data, feed speed data, pipe wall deformation feedback data and spindle vibration data of the target cutting machine.
[0016] The drive motor current data, feed speed data, pipe wall deformation feedback data, and spindle vibration data are subjected to amplitude mapping to obtain the drive motor current signal, feed speed signal, pipe wall deformation feedback signal, and spindle vibration signal of the target cutting machine.
[0017] In a preferred embodiment, the step of performing component analysis on the drive motor current signal to obtain the reference load of the target cutting machine includes:
[0018] The drive motor current signal is decomposed hierarchically to obtain the dimensional component data of the target cutting machine;
[0019] The effective load of the target cutting machine is obtained by performing steady-state screening on the scale component data;
[0020] Based on the effective load, the amplitude of the scale component data is merged to obtain the reference load of the target cutting machine.
[0021] In a preferred embodiment, the step of performing condition-related calibration on the reference load based on the feed speed signal and the spindle vibration signal to obtain the corrected load of the target cutting machine includes:
[0022] The feed speed signal and the spindle vibration signal are identified to be in the same direction to obtain the directional scale of the target cutting machine;
[0023] Based on the directional scale, the difference between the feed speed signal and the spindle vibration signal is measured to obtain the deviation criterion of the target cutting machine;
[0024] Based on the directional scale and the deviation criterion, the feed speed signal and the spindle vibration signal are jointly determined to obtain the operating attributes of the target cutting machine;
[0025] Based on the operating attributes, the baseline load is corrected to obtain the corrected load for the target cutting machine.
[0026] In a preferred embodiment, the step of sensing the difference trend between the corrected load and the reference load to obtain the load deviation trend of the target cutting machine includes:
[0027] The deviation component of the target cutting machine is obtained by decomposing the difference between the modified load and the reference load.
[0028] The deviation component is morphologically segmented to obtain the variation segments of the deviation component;
[0029] By performing correlation evolution analysis on the changed segments, the load deviation trend of the target cutting machine is obtained.
[0030] In a preferred embodiment, the step of performing correlation evolution analysis on the changed segments to obtain the load deviation trend of the target cutting machine includes:
[0031] Feature extraction is performed on the changed segment to obtain the morphological feature information of the changed segment;
[0032] Based on the morphological feature information, similarity clustering is performed on the changed segments to obtain the classification representation of the changed segments;
[0033] Based on the classification representation, the temporal sequence of the changed segments is deduced to obtain the connection relationship of the changed segments;
[0034] Based on the aforementioned connection relationship, the trajectory of the changing segment is reconstructed to obtain the load deviation trend of the target cutting machine.
[0035] In a preferred embodiment, the step of performing deformation-load collaborative analysis on the load deviation trend based on the pipe wall deformation feedback signal to obtain the comprehensive risk deviation of the target cutting machine includes:
[0036] The load amplitude of the target cutting machine is obtained by parametric analysis of the load deviation trend.
[0037] Numerical analysis is performed on the pipe wall deformation feedback signal to obtain the deformation amplitude of the target cutting machine;
[0038] The load amplitude and the deformation amplitude are coupled and measured to obtain the comprehensive risk deviation of the target cutting machine. The formula for calculating the comprehensive risk deviation is as follows:
[0039] ;
[0040] in, This indicates the overall risk deviation. This indicates the load amplitude. This indicates the deformation amplitude. This represents the preset deformation matching coefficient. This represents the maximum deviation amplitude of the load deviation trend across the entire time domain. This represents the natural exponential function.
[0041] In a preferred embodiment, the step of performing deviation inversion on the benchmark load based on the load deviation trend to obtain the risk assessment range of the target cutting machine includes:
[0042] Deviation sampling is performed on the load deviation trend to obtain the deviation distribution of the target cutting machine;
[0043] Based on the reference load, a reference offset mapping is performed on the deviation distribution to obtain the offset range of the target cutting machine;
[0044] The risk assessment range of the target cutting machine is obtained by calibrating the boundary of the offset range.
[0045] In a preferred embodiment, the step of performing a reference offset mapping on the deviation distribution based on the reference load to obtain the offset range of the target cutting machine includes:
[0046] The deviation values in the deviation distribution are time-series anchored to obtain the time-series position of the deviation of the target cutting machine;
[0047] Based on the deviation time sequence position, the amplitude of the reference load is sampled to obtain the deviation amplitude sequence of the reference load;
[0048] Extreme value picking is performed on the deviation amplitude sequence to obtain the maximum and minimum deviation amplitudes of the target cutting machine;
[0049] Based on the maximum deviation amplitude and the minimum deviation amplitude, the boundary of the reference load is defined to obtain the offset range of the target cutting machine.
[0050] In a preferred embodiment, the step of predicting the risk situation based on the comprehensive risk deviation degree according to the risk assessment interval to obtain the risk level of the target cutting machine includes:
[0051] Based on the risk assessment interval, the location of the comprehensive risk deviation is identified to obtain the interval relative position of the comprehensive risk deviation;
[0052] Based on the relative position within the interval, the trend of the comprehensive risk deviation is extrapolated to obtain the risk trend of the target cutting machine;
[0053] The risk trends are summarized and integrated to obtain the risk situation of the target cutting machine;
[0054] Based on a preset risk level library, the risk situation is hierarchically matched to obtain the risk level of the target cutting machine.
[0055] Compared with the prior art, the present invention has the following beneficial effects:
[0056] 1. This invention, by synchronously collecting multi-dimensional sensing signals such as motor current, feed speed, pipe wall deformation, and spindle vibration, and through layered analysis, working condition calibration, and trend extrapolation coupled calculation of comprehensive risk deviation, can completely capture the linkage changes of load, pipe deformation, and equipment vibration during the operation of the cutting machine. This significantly improves the completeness and accuracy of identifying abnormal equipment operation characteristics, shortens the risk identification response time, and improves the overall efficiency of the cutting machine operation risk early warning work.
[0057] 2. This invention forms a dedicated risk assessment range by relying on load deviation trend inversion, and combines quantitative formulas to achieve collaborative analysis of load and pipe wall deformation data. Based on the range position, it completes the risk situation stratification assessment, which can finely distinguish different levels of equipment operation hazards and accurately output the corresponding risk level. It provides a quantitative and intuitive judgment basis for the pre-treatment of cutting machine failures, and effectively reduces the probability of various operational failures such as pipe breakage, equipment overload, and spindle wear during the cutting operation. Attached Figure Description
[0058] Figure 1 A flowchart illustrating a method for early warning of operational risks in cutting cable protection pipes, provided in an embodiment of the present invention;
[0059] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0060] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0061] This application provides a method for early warning of operational risks of a cutting machine used for cutting cable protection pipes. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for early warning of operational risks of a cutting machine used for cutting cable protection pipes can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.
[0062] Reference Figure 1 The diagram shown is a flowchart illustrating a method for early warning of operational risks in cutting cable protection pipes according to an embodiment of the present invention. In this embodiment, the method for early warning of operational risks in cutting cable protection pipes includes:
[0063] P1. Acquire the drive motor current signal, feed speed signal, pipe wall deformation feedback signal and spindle vibration signal of the target cutting machine, and perform component analysis on the drive motor current signal to obtain the reference load of the target cutting machine.
[0064] In this embodiment of the invention, acquiring the drive motor current signal, feed speed signal, pipe wall deformation feedback signal, and spindle vibration signal of the target cutting machine includes:
[0065] Collect raw sensor data from the target cutting machine;
[0066] The original sensor data is filtered to remove interference, resulting in the denoised sampling data of the target cutting machine.
[0067] The denoised sampling data is sorted to obtain the drive motor current data, feed speed data, pipe wall deformation feedback data and spindle vibration data of the target cutting machine.
[0068] The drive motor current data, feed speed data, pipe wall deformation feedback data, and spindle vibration data are subjected to amplitude mapping to obtain the drive motor current signal, feed speed signal, pipe wall deformation feedback signal, and spindle vibration signal of the target cutting machine.
[0069] The step of performing component analysis on the drive motor current signal to obtain the reference load of the target cutting machine includes:
[0070] The drive motor current signal is decomposed hierarchically to obtain the dimensional component data of the target cutting machine;
[0071] The effective load of the target cutting machine is obtained by performing steady-state screening on the scale component data;
[0072] Based on the effective load, the amplitude of the scale component data is merged to obtain the reference load of the target cutting machine.
[0073] Dedicated signal transmission paths were established to connect to the target cutting machine's drive motor sensing module, feed speed detection module, pipe wall deformation detection module, and spindle vibration sensing module. These paths continuously and synchronously received electrical data output from each sensing module in real time. Information from all sensing modules within the same acquisition cycle was uniformly collected and summarized to form the target cutting machine's raw sensor data. This data was then thoroughly investigated for any abnormal fluctuations caused by electromagnetic interference, accidental mechanical impacts, or unstable wiring. Each piece of information was individually identified and all interfering information was removed, ensuring a complete and objective record reflecting the true performance of the cutting machine. Effective information on the operating conditions is used to obtain denoised sampling data for the target cutting machine. The monitoring object corresponding to each piece of information in the denoised sampling data is identified strictly according to the pre-set collection source classification rules. All information from the drive motor monitoring module, all information representing the operating speed of the feed mechanism, all information generated by the feedback from the pipe wall deformation detection device, and all information collected from the spindle vibration monitoring device are integrated and collected. After systematic classification and sorting, the drive motor current data, feed speed data, pipe wall deformation feedback data, and spindle vibration data of the target cutting machine are obtained.
[0074] The standard signal range corresponding to the output values of various sensing components after the calibration work was completed was retrieved. All the acquisition records stored in the internal storage of drive motor current data, feed speed data, pipe wall deformation feedback data and spindle vibration data were completely read. Standardization conversion was performed on all values contained in each type of data. The original output formats of different sensor models were uniformly adjusted to the pre-agreed standard signal expression form. This eliminated the data docking obstacles caused by the inconsistency of output specifications of various sensing devices, and ensured that subsequent multi-channel signals could be synchronously analyzed and coordinated. After all categories of data were standardized and converted, the drive motor current signal, feed speed signal, pipe wall deformation feedback signal and spindle vibration signal of the target cutting machine were obtained.
[0075] Based on the various information within the drive motor current signal and their corresponding equipment operation impact levels, a complete layer-by-layer decomposition process is carried out. The various information components corresponding to continuous cutting conditions, equipment start-up and shutdown instantaneous conditions, and no-load stable operation conditions are stripped away and distinguished layer by layer. The independent information segments corresponding to different equipment operation behaviors are clearly divided. All the subdivided information segments generated during the entire decomposition process are completely saved and arranged in an orderly manner according to the acquisition time sequence. A complete data set with a multi-layered progressive information structure is constructed. After the overall decomposition of the drive motor current signal is completed, the dimensional component data of the target cutting machine is obtained.
[0076] For each set of subdivided information contained in the scale component data, the working condition persistence feature identification work is continuously carried out, and the actual persistence duration corresponding to each set of component information is continuously tracked and recorded. From this, information content that exists stably for a long time and can continuously correspond to the stable cutting operation of the cutting machine is selected, and information content with short persistence duration, only corresponding to the instantaneous action of the equipment or random fluctuations is removed. Only data segments that can stably follow the changing law of cutting load are retained. After performing all the filtering and differentiation operations and removing all invalid fluctuation information, the effective load of the target cutting machine is obtained.
[0077] Using the effective load obtained after screening as a unified reference benchmark, the magnitude of each type of effective information retained within the scale component data is adjusted to standardize and unify the representation scale of information collected in different working periods. This eliminates the lateral comparison error that is easily caused by the inconsistency of data magnitude in different cutting stages. The integration and merging of all effective information is carried out in an orderly manner, so that the integrated data can objectively and evenly represent the normal cutting load level of the cutting machine. After the entire amplitude merging process is completed, the reference load of the target cutting machine is obtained.
[0078] The beneficial effects include the simultaneous acquisition of multi-dimensional operational sensor information of the cutting machine, followed by interference filtering, data sorting, and amplitude mapping processing. This unifies the signal expression standards of multi-channel monitoring data, ensuring that the acquired information accurately reflects the real-time operating status of the equipment. For the drive motor current signal, hierarchical decomposition, steady-state screening, and amplitude merging processing are performed progressively. This effectively separates invalid data generated by brief disturbances, stably extracts the benchmark load that can characterize conventional cutting conditions, and reduces the adverse effects of random fluctuations on the load determination results. This provides highly consistent and reliable basic analysis data for subsequent comprehensive evaluation of the cutting machine's cutting operation status based on various monitoring signals.
[0079] P2. Based on the feed speed signal and the spindle vibration signal, the reference load is calibrated according to the working condition to obtain the corrected load of the target cutting machine;
[0080] In this embodiment of the invention, the step of performing condition-related calibration on the reference load based on the feed speed signal and the spindle vibration signal to obtain the corrected load of the target cutting machine includes:
[0081] The feed speed signal and the spindle vibration signal are identified to be in the same direction to obtain the directional scale of the target cutting machine;
[0082] Based on the directional scale, the difference between the feed speed signal and the spindle vibration signal is measured to obtain the deviation criterion of the target cutting machine;
[0083] Based on the directional scale and the deviation criterion, the feed speed signal and the spindle vibration signal are jointly determined to obtain the operating attributes of the target cutting machine;
[0084] Based on the operating attributes, the baseline load is corrected to obtain the corrected load for the target cutting machine.
[0085] According to the unified acquisition time sequence, the feed speed signal and the spindle vibration signal are precisely aligned to cover all time sequence sampling nodes. The dynamic fluctuations of the two types of signals at each time node are continuously tracked, and the corresponding changes and development trends are continuously and completely recorded. Various linkage manifestations of synchronous rise or fall of the two types of signals are continuously collected. All feature information corresponding to the consistent trend of the two types of signals in all time periods within the complete cutting operation cycle is collected. The linkage features collected throughout the process are uniformly collected and standardized. The inherent laws of the linkage changes between the changes in the motion state of the feed mechanism and the changes in the vibration state of the spindle are completely sorted out. All the integrated linkage feature information is stored and filed in a unified standard format. After the entire set of same-direction identification operations is completed, the direction scale of the target cutting machine is obtained.
[0086] Based on the standard linkage relationship between the feed speed signal and the spindle vibration signal recorded within the already fully generated directional scale, the entire process is compared against the reference benchmark. The changes in the feed speed signal and the spindle vibration signal in real time are sequentially traversed across all time-series sampling nodes. The location where the actual signal change deviates from the predetermined standard linkage law is identified one by one. All signal differences that occur during the entire operation cycle are fully summarized. All the collected difference features are systematically classified and sorted, and a complete and comparable feature reference system is built. The signal difference features corresponding to different working conditions are uniformly collected and integrated to form a standardized set of discrimination criteria. After the entire set of difference measurement operations is completed, the deviation criteria of the target cutting machine are obtained.
[0087] The system retrieves the signal standard linkage characteristics recorded in the pre-processed directional scale and the various signal deviation characteristics summarized by the deviation criteria. Under a unified time sequence dimension, it continuously compares the dynamic data of the feed speed signal and the spindle vibration signal in real time. It continuously compares the degree of fit between the real-time signal combination state and the standard linkage characteristics. Based on the different signal matching performance, it accurately distinguishes various non-overlapping working conditions such as continuous cutting operation, no-load feed movement, and abnormal spindle vibration. It continuously and qualitatively identifies the specific working condition of the cutting machine and completely retains the identification results at each moment. It integrates the continuous working condition identification information of the whole time period to complete the systematic working condition classification. After the entire set of collaborative judgment operations is completed, the operating attributes of the target cutting machine are obtained.
[0088] Based on the real-time operating conditions of the cutting machine corresponding to the operating attributes obtained in the collaborative judgment step, the pre-set and mutually compatible load correction specifications are retrieved. For the reference load obtained by analyzing the drive motor current signal layer by layer in the previous processing step, the correction and adjustment processing is carried out to match the real-time operating conditions of the equipment. This specifically compensates for the data deviation problem caused by the reference load being generated only based on current information and not incorporating the coupling effect of feed motion state and spindle vibration state. This allows the adjusted load information to synchronously and objectively reflect the comprehensive effect of feed operation state and spindle vibration on cutting load. The working condition adaptation and correction work of each data content within the reference load is continuously completed item by item. After the entire load correction operation is fully executed, the corrected load of the target cutting machine is obtained.
[0089] The beneficial effects are that the feed speed signal and spindle vibration signal are sequentially identified and measured in the same direction, and then processed collaboratively to fully explore the coupling relationship between the feed action and the spindle vibration, accurately identify the current operating attributes of the cutting machine, and perform working condition correlation calibration on the benchmark load based on the actual operating attributes of the equipment. This improves the shortcomings of relying solely on motor current to solve the load, which cannot take into account the continuous changes in dynamic cutting conditions. It effectively reduces the load judgment deviation caused by working condition fluctuations, and ensures that the final corrected load can objectively and completely reflect the actual cutting load level of the cutting machine. This provides more accurate and more adaptable basic analysis data for subsequent monitoring of the entire cutting process and identification and early warning of abnormal working conditions.
[0090] P3. Sensing the difference trend between the corrected load and the reference load to obtain the load deviation trend of the target cutting machine;
[0091] In this embodiment of the invention, the step of sensing the difference trend between the corrected load and the reference load to obtain the load deviation trend of the target cutting machine includes:
[0092] The deviation component of the target cutting machine is obtained by decomposing the difference between the modified load and the reference load.
[0093] The deviation component is morphologically segmented to obtain the variation segments of the deviation component;
[0094] By performing correlation evolution analysis on the changed segments, the load deviation trend of the target cutting machine is obtained.
[0095] The correlation evolution analysis of the changed segments to obtain the load deviation trend of the target cutting machine includes:
[0096] Feature extraction is performed on the changed segment to obtain the morphological feature information of the changed segment;
[0097] Based on the morphological feature information, similarity clustering is performed on the changed segments to obtain the classification representation of the changed segments;
[0098] Based on the classification representation, the temporal sequence of the changed segments is deduced to obtain the connection relationship of the changed segments;
[0099] Based on the aforementioned connection relationship, the trajectory of the changing segment is reconstructed to obtain the load deviation trend of the target cutting machine.
[0100] The complete records of the corrected load and the reference load, which are matched one-to-one at each time point within the complete cutting cycle, are retrieved. The load values of the two sets of load data are compared item by item along a unified and continuous time axis at each acquisition time. All objective differences between the two sets of load data are broken down layer by layer. The various independent information units introduced from the feed speed signal and spindle vibration signal working condition correlation calibration are completely distinguished. All independent difference information units generated during the difference deconstruction process are continuously collected. All collected difference information units are organized and stored in an orderly manner according to the original acquisition time sequence to ensure that the difference information retains the corresponding time attributes. After the entire difference deconstruction operation is completed, the deviation components of the target cutting machine are obtained.
[0101] By continuously observing the dynamic changes of the deviation component throughout the entire timeline of the cutting operation, the continuous change curve is formed. The transition boundary nodes where the change direction and fluctuation mode of all changes within the curve change significantly are accurately identified. All identified transition boundary nodes are used as fixed dividing boundaries to perform segmentation processing on the entire continuous change curve. The originally continuous and complete deviation change process is divided into multiple independent segments, each with independent fluctuation characteristics and independent duration. Each segment is stored separately after the segmentation is completed, and the start and end time marker information of each segment is fully preserved to ensure that each segment can be traced back to the corresponding operation period. After the entire morphological segmentation operation is completed, the change fragment of the deviation component is obtained.
[0102] All segmented change fragments are retrieved sequentially for feature extraction. This process fully identifies various external characteristics of a single change fragment, including its continuous fluctuations, duration, inflection points, and state transitions. All identification information for each change fragment is compiled and a corresponding feature file is created to form morphological feature information. Based on the extracted morphological feature information, the matching of change patterns between different change fragments is compared horizontally segment by segment. Change fragments with consistent fluctuation patterns are grouped into the same group and assigned a unified category label to achieve similarity clustering and classification representation. Based on the classification representation of each group, the arrangement patterns of different types of change fragments appearing sequentially are continuously analyzed, and various combinations of fragments are recorded to obtain the connection relationships between the change fragments.
[0103] The established relationships between different segments are used as a reference for trajectory splicing. The time sequence of the original sensor data acquisition is strictly followed. The sequential combination rules obtained through time-series deduction between different segment categories are used to systematically splice and integrate all the scattered and independent segments. The segments that were originally separated and stored are connected end to end according to the complete evolution logic. The deviation information is fully restored as it evolves continuously with the cutting operation. The complete trajectory of the load deviation from the start to the end of the operation is continuously restored. After the entire trajectory reconstruction operation is completed, the load deviation trend of the target cutting machine is obtained.
[0104] The beneficial effects are as follows: First, differential decomposition and morphological segmentation operations are performed on the corrected load and the reference load in sequence to decompose the overall load difference into change segments with independent characteristics. Then, feature extraction, similarity clustering, time series deduction, and trajectory reconstruction are carried out in a continuous multi-level process. This can accurately restore the complete development path of the load deviation as it evolves with the cutting operation, fully retain the various stage-by-stage subtle evolution characteristics during the load change process, avoid the defects of losing local short-term fluctuation information caused by directly comparing the overall data, and clearly present the dynamic evolution law of the load deviation caused by the working condition calibration process. This provides detailed and time-series-continuous logical data support for subsequent evaluation of the gradual trend of the cutting state of the cutting machine and early identification of slow drift phenomena in equipment operation.
[0105] P4. Based on the pipe wall deformation feedback signal, perform deformation load collaborative analysis on the load deviation trend to obtain the comprehensive risk deviation of the target cutting machine;
[0106] In this embodiment of the invention, the step of performing deformation-load collaborative analysis on the load deviation trend based on the pipe wall deformation feedback signal to obtain the comprehensive risk deviation of the target cutting machine includes:
[0107] The load amplitude of the target cutting machine is obtained by parametric analysis of the load deviation trend.
[0108] Numerical analysis is performed on the pipe wall deformation feedback signal to obtain the deformation amplitude of the target cutting machine;
[0109] The load amplitude and the deformation amplitude are coupled and measured to obtain the comprehensive risk deviation of the target cutting machine. The formula for calculating the comprehensive risk deviation is as follows:
[0110] ;
[0111] in, This indicates the overall risk deviation. This indicates the load amplitude. This indicates the deformation amplitude. This represents the preset deformation matching coefficient. This represents the maximum deviation amplitude of the load deviation trend across the entire time domain. This represents the natural exponential function.
[0112] The entire continuous evolution trajectory information of the load deviation trend is fully traversed across the entire time domain. The dynamic changes of the deviation in each continuous period of the trajectory are read sequentially along the time axis corresponding to the continuous progress of the cutting operation. Irrelevant information in the load deviation trend that is only used to mark the time sequence and does not represent the intensity of fluctuation is filtered out. The core feature information that can objectively represent the strength of the deviation fluctuation in the entire continuous trajectory is continuously captured. The fluctuation magnitude related content corresponding to different processing periods is fully collected and uniformly organized. All effective feature information that can intuitively reflect the change law of load deviation fluctuation intensity is retained throughout the process. No reference-worthy magnitude features are omitted in each stage of the trajectory. After the entire set of parameter extraction operations is completed, the load amplitude of the target cutting machine is obtained.
[0113] The system continuously reads all the time-series acquisition information of the pipe wall deformation feedback signal during the entire cutting operation cycle. It distinguishes layer by layer the secondary fluctuations caused by external environmental disturbances and slight vibrations of the detection element in the pipe wall deformation feedback signal, as well as the main changes corresponding to the deformation of the pipe caused by the cutting extrusion force. It completely eliminates all kinds of interference information that cannot represent the true deformation state of the pipe. It systematically sorts out the magnitude change information corresponding to the continuous evolution of the pipe deformation degree as the cutting process progresses. It centrally collects the core feature content that can intuitively represent the strength of the pipe deformation and organizes it according to the original acquisition time sequence. It completely preserves the dynamic change characteristics of the deformation as the processing progresses. After the entire set of numerical analysis operations is completed, the deformation amplitude of the target cutting machine is obtained.
[0114] The deformation matching coefficients, which have been pre-calibrated and fully stored through multiple operating condition tests in the early stage of the system, are retrieved. At the same time, the maximum deviation amplitude of the load deviation trend in the entire time domain, which is obtained by statistically summarizing the entire load deviation trend in the early stage, is retrieved. The load amplitude output by the parameter extraction stage and the deformation amplitude output by the numerical analysis stage are used as two basic feature data required for the coupled measurement operation. The basic correlation relationship between the load amplitude and the deformation amplitude is established. Based on the deformation matching coefficients, the deformation matching degree of the established basic correlation results is corrected, and the basic measurement subject that links the two types of amplitudes is established.
[0115] The load amplitude and the maximum deviation amplitude in the entire time domain are further compared and constrained. The correlation results obtained after deformation matching coefficient correction are further calibrated in numerical range. According to the established logic, each progressive processing link is connected layer by layer to form a unified and complete quantitative index. The entire process of multi-level progressive correlation comparison, coefficient correction and magnitude calibration is completed. The processing results generated at all levels are integrated to form the final evaluation index. After the entire set of coupled measurement operations are completed, the comprehensive risk deviation of the target cutting machine is obtained.
[0116] After collecting raw sensor data from the target cutting machine, interference filtering and data sorting are performed. Amplitude mapping is then executed to generate drive motor current signals, feed speed signals, pipe wall deformation feedback signals, and spindle vibration signals. The drive motor current signal is decomposed hierarchically to obtain dimensional component data. After steady-state screening, an effective load is generated, and amplitude merging is performed to generate a reference load. Based on the feed speed signal and spindle vibration signal, unidirectional identification is performed to obtain the direction scale. Difference measurement is then used to generate a deviation criterion. Combining the direction scale deviation criterion, the operating attributes are collaboratively determined, and the reference load is corrected based on the operating attributes to generate a corrected load. The difference between the corrected load and the reference load is then deconstructed and generated. The deviation component is morphologically segmented to obtain change segments. The morphological feature information of the change segments is extracted and classified through similarity clustering. The time series deduction is used to obtain the connection relationship trajectory reconstruction to produce the load deviation trend. The load deviation trend parameter is directly extracted to produce the load amplitude. The pipe wall deformation feedback signal is numerically analyzed to directly produce the deformation amplitude. The deformation matching coefficient is a pre-set fixed value. After extracting all values in the entire time domain of the load deviation trend, the maximum value is selected to produce the maximum deviation amplitude. The natural exponential function is based on fixed mathematical operation rules. All the above outputs are used as all the basic materials required to calculate the comprehensive risk deviation degree.
[0117] The load amplitude corresponds to the real-time fluctuation of the motor load during the cutting process of the cutting machine, and the deformation amplitude corresponds to the real-time deformation of the cable protection pipe caused by cutting and squeezing. The first-level calculation result is obtained by dividing the load amplitude by the deformation amplitude and subtracting a fixed value. The difference between the load amplitude and the deformation amplitude is extracted, divided by the maximum deviation amplitude, and multiplied by the deformation matching coefficient to obtain the second-level calculation result. The second-level calculation result is input into the natural exponential function to obtain the third-level calculation result. The first-level calculation result and the third-level calculation result are multiplied to finally produce the comprehensive risk deviation degree. The comprehensive risk deviation degree is used to synchronously quantify the overall deviation degree after the superposition of two abnormal factors, namely motor load fluctuation and pipe extrusion deformation. The comprehensive risk deviation degree value directly serves as the core quantitative basis for predicting the risk level in the matching risk judgment interval, realizing the collaborative quantitative analysis of two types of data, namely pipe wall deformation feedback signal and load deviation trend, to eliminate the problem of missed and false alarms caused by single sensor parameter judgment.
[0118] When the load amplitude is greater than the deformation amplitude, the result of the first-level calculation is positive. The value of the second-level calculation increases synchronously with the widening difference between the load amplitude and the deformation amplitude. The output value of the natural exponential function also increases synchronously. After multiplying the results of the two calculations, the comprehensive risk deviation continues to increase positively. When the load amplitude is equal to the deformation amplitude, the value of the first-level calculation is zero. The comprehensive risk deviation is fixed at zero, indicating that the cutting machine is in a standard, deviation-free operating state. When the load amplitude is less than the deformation amplitude, the result of the first-level calculation is negative. The value of the second-level calculation decreases synchronously with the widening difference between the deformation amplitude and the load amplitude. The output value of the natural exponential function also decreases synchronously. After multiplying the results of the two calculations, the comprehensive risk deviation continues to decrease negatively. The larger the positive value of the comprehensive risk deviation, the greater the degree of abnormal motor load compared to the degree of abnormal pipe deformation, indicating a continuous increase in the risk of equipment overload. The smaller the negative value of the comprehensive risk deviation, the greater the degree of abnormal pipe extrusion deformation compared to the degree of abnormal motor load, indicating a continuous increase in the risk of pipe edge breakage and scrapping. The larger the deviation of the comprehensive risk deviation from zero, the higher the risk level obtained after subsequent matching of the risk judgment interval.
[0119] The beneficial effects are as follows: first, the load amplitude is extracted from the load deviation trend, and the deformation amplitude is analyzed from the pipe wall deformation feedback signal. Then, based on the preset matching coefficient and the maximum deviation amplitude in the time domain, a multi-level coupling measurement of the two types of amplitudes is completed. Simultaneously, the monitoring information of two independent dimensions, namely the cutting load deviation evolution trend and the real-time deformation state of the pipe, is integrated. This avoids the one-sided evaluation problem caused by risk assessment based on single-dimensional monitoring data, and realizes the collaborative comprehensive assessment of equipment load conditions and workpiece deformation state. This allows the comprehensive risk deviation degree to fully reflect the various abnormal tendencies hidden in the entire pipe cutting process, and significantly improves the completeness, objectivity and reliability of the risk assessment results of pipe cutting conditions.
[0120] P5. Based on the load deviation trend, perform deviation inversion on the benchmark load to obtain the risk judgment range of the target cutting machine;
[0121] In this embodiment of the invention, the step of performing deviation inversion on the benchmark load based on the load deviation trend to obtain the risk assessment range of the target cutting machine includes:
[0122] Deviation sampling is performed on the load deviation trend to obtain the deviation distribution of the target cutting machine;
[0123] Based on the reference load, a reference offset mapping is performed on the deviation distribution to obtain the offset range of the target cutting machine;
[0124] The risk assessment range of the target cutting machine is obtained by calibrating the boundary of the offset range.
[0125] The step of performing a reference offset mapping on the deviation distribution based on the reference load to obtain the offset range of the target cutting machine includes:
[0126] The deviation values in the deviation distribution are time-series anchored to obtain the time-series position of the deviation of the target cutting machine;
[0127] Based on the deviation time sequence position, the amplitude of the reference load is sampled to obtain the deviation amplitude sequence of the reference load;
[0128] Extreme value picking is performed on the deviation amplitude sequence to obtain the maximum and minimum deviation amplitudes of the target cutting machine;
[0129] Based on the maximum deviation amplitude and the minimum deviation amplitude, the boundary of the reference load is defined to obtain the offset range of the target cutting machine.
[0130] The load deviation trend is retrieved by continuously recording the dynamic changes of load deviation within the complete processing cycle of the target cutting machine. Uninterrupted, full-coverage deviation sampling is conducted strictly along the complete timeline formed by the continuous operation of the equipment. All original load deviation records corresponding to each subdivided time node within the load deviation trend recording cycle are extracted sequentially according to equal and continuous time intervals. This fully covers the deviation data corresponding to all operating conditions of the equipment during no-load trial operation, material cutting, processing, standby, and buffering, without skipping any continuous running segment. All collected original deviation records are strictly collected and summarized according to the chronological order of equipment operation. All sampled original deviation information is then systematically arranged to form a complete deviation distribution of the target cutting machine that possesses continuous temporal characteristics and covers multiple operating conditions.
[0131] The process involves retrieving the complete target cutting machine deviation distribution from the sampled dataset, extracting each independent deviation value within the distribution, and identifying the corresponding equipment operating time node for each deviation value. An immutable binding relationship is established between each deviation value and its corresponding operating time node. This process is repeated for all deviation values and their associated operating times, and all bound time-series information is systematically organized. This generates a precise time-series position for the target cutting machine, accurately recording the time point corresponding to each deviation. The process then retrieves the target cutting machine's deviation time-series position after all time-series association information has been solidified. Based on the precise time points recorded in the deviation time-series position, the load values corresponding to the same operating time within the baseline load dataset are retrieved sequentially. The baseline load values matching each time-series point are collected and arranged in an orderly manner, strictly following the chronological order of operating time. This process generates a complete time-series position for each deviation occurrence. A corresponding baseline load deviation amplitude sequence is obtained. The complete time-series arrangement of the baseline load deviation amplitude sequence is retrieved. All load amplitude records stored in the sequence are traversed from beginning to end for comparison and screening. Records at the upper and lower limits of the values are extracted from all amplitude records. The two sets of filtered amplitude contents are set as the maximum and minimum deviation amplitudes of the target cutting machine, respectively. The maximum and minimum deviation amplitudes of the target cutting machine with known numerical contents are retrieved. The standard operating value corresponding to the baseline load is used as the center reference of the interval. The upper limit boundary of the interval is determined by extending outward based on the maximum deviation amplitude, and the lower limit boundary of the interval is determined by extending outward based on the minimum deviation amplitude. After all the boundary points of the upper and lower sets are defined, the complete numerical interval enclosed by the two sets of boundaries is integrated to generate the offset range of the target cutting machine that is adapted to the load fluctuation characteristics of the equipment.
[0132] The offset range of the target cutting machine, obtained by fully mapping the baseline load, is retrieved. The pre-set interval hierarchical division specification for cutting machine load risk identification is retrieved. Fine-grained boundary calibration is carried out for the upper and lower boundaries of the offset range and the transition positions within the interval. The operational risk level boundary points corresponding to each continuous segment within the offset range are clearly identified. The division boundary between the normally stable operating segment and the risk segment with potential operational hazards within the interval is clearly distinguished. All confirmed boundary points are uniformly registered and marked. The interval content with complete hierarchical boundary markings is integrated and summarized to generate a complete risk judgment interval for the target cutting machine that can be directly used for real-time operation status judgment and has a clear hierarchical division.
[0133] The beneficial effects are as follows: First, full-coverage deviation sampling is performed along the complete operating sequence to collect deviation data under multiple operating conditions throughout the entire cycle, constructing a deviation distribution that can objectively reflect the historical deviation characteristics of the equipment. Then, time-series anchoring is performed on all deviation values within the deviation distribution to establish the correspondence between deviation and operating time. Next, based on time-series points, the baseline load amplitude is picked up to construct a deviation amplitude sequence. The maximum and minimum deviation amplitudes are obtained through traversal and filtering to complete the delineation of the offset range. Finally, the offset range is finely delineated to generate a risk judgment interval. The entire hierarchical processing process is bound to the time-series correlation characteristics of equipment load changes, abandoning the processing method of simple static numerical calculations. All intermediate output data are interconnected and the information is complete and can be traced and verified independently. It can complete the interval construction in accordance with the load fluctuation characteristics under the dynamic processing conditions of the cutting machine, effectively eliminating the problem of risk interval delineation being out of sync with actual operating conditions due to ignoring time-series characteristics, and significantly improving the accuracy of subsequent cutting machine load anomaly monitoring and risk identification.
[0134] P6. Based on the risk assessment interval, perform risk situation prediction on the comprehensive risk deviation to obtain the risk level of the target cutting machine.
[0135] In this embodiment of the invention, the step of predicting the risk situation based on the comprehensive risk deviation degree according to the risk determination interval to obtain the risk level of the target cutting machine includes:
[0136] Based on the risk assessment interval, the location of the comprehensive risk deviation is identified to obtain the interval relative position of the comprehensive risk deviation;
[0137] Based on the relative position within the interval, the trend of the comprehensive risk deviation is extrapolated to obtain the risk trend of the target cutting machine;
[0138] The risk trends are summarized and integrated to obtain the risk situation of the target cutting machine;
[0139] Based on a preset risk level library, the risk situation is hierarchically matched to obtain the risk level of the target cutting machine.
[0140] Retrieve the risk assessment range of the target cutting machine, which has been fully calibrated through the aforementioned steps, and the comprehensive risk deviation obtained through real-time data acquisition and calculation. Incorporate the complete numerical information corresponding to the comprehensive risk deviation into the full range of values defined by the risk assessment range to conduct refined location identification. Verify the location against the boundaries of all pre-defined sub-segments within the risk assessment range to accurately confirm that the comprehensive risk deviation currently falls within the corresponding exclusive sub-segment of the risk assessment range. Completely record the spatial location information corresponding to the comprehensive risk deviation from the upper and lower boundaries of the range and the risk boundary points at all levels. Summarize all location information and organize and archive it in an orderly manner to generate the relative position of the comprehensive risk deviation within the range system, which can accurately depict the location of the current deviation value.
[0141] The system retrieves the relative position of the comprehensive risk deviation from the location identification, continuously collects historical data of the relative position of the intervals output sequentially at multiple consecutive equipment operation sampling times, strictly follows the time sequence of the target cutting machine's continuous processing operation to sort out the entire process of the continuous displacement changes of the relative position of the intervals, continuously tracks the evolution path and continuous evolution law of the relative position of the intervals towards the boundary points of various sections within the risk judgment interval, fully sorts out the positional change and development trend of the comprehensive risk deviation in the subsequent continuous processing period, comprehensively collects the positional evolution details formed by different time nodes, and fully generates the risk trend of the target cutting machine that can clearly characterize the subsequent dynamic evolution of the risk status of the target cutting machine.
[0142] The risk trends of all target cutting machines obtained from segmented deduction under continuous time sequence are retrieved. The independent risk trend content generated from different sampling periods is collected in strict accordance with the time sequence of the cutting machine's processing and operation. The temporal continuity relationship and state transition characteristics between risk trends in adjacent time periods are uniformly sorted out. Scattered instantaneous transition information without continuous reference value generated during the time sequence connection process is filtered out. Multiple segmented risk trends with mutual evolution and inheritance relationship are integrated into a coherent and complete state evolution context. After the collection, sorting and connection of all segmented trend information are completed, they are uniformly summarized and integrated to form a complete risk situation of the target cutting machine covering the entire risk evolution process within a complete continuous processing cycle.
[0143] The risk situation of the target cutting machine, which has been integrated according to the timeline, and the risk level library, which has been pre-built and archived in a standardized manner, are retrieved. The risk level library contains a complete collection of risk situation samples with different evolutionary characteristics and their corresponding exclusive risk level labels. All feature information contained in the risk situation, such as evolution trend, duration of stay, migration characteristics, and state transition nodes, is extracted one by one. The complete feature elements of the risk situation are compared and matched with each set of preset situation samples stored in the risk level library. The exclusive level label content that can be completely matched with the feature elements is identified. The matched level label is used as the current risk assessment conclusion of the equipment, and the risk level of the target cutting machine after standardization and classification is generated.
[0144] The beneficial effects are as follows: relying on the risk assessment interval, the comprehensive risk deviation is refined to identify the location information of the real-time value within the interval system, and the relative position of the interval is obtained. Based on the relative position of the interval in continuous time series, dynamic changes are sorted out to complete the trend deduction and obtain the segmented risk trend. Then, multiple time series related risk trends are connected and redundant information is cleaned up and integrated to form a complete risk situation. Finally, a standardized risk level library is called to achieve accurate matching between the risk situation and the evaluation level and output the risk level of the target cutting machine. The whole process no longer relies solely on the static risk assessment based on the value at a single moment. It simultaneously integrates the current location information and the long-term evolution and development law. Each processing step is progressive and the data carriers are completely connected. All operation records can be retained and verified. It fully adapts to the dynamic fluctuation of the load during the continuous processing of the cutting machine, effectively avoids the deviation of the risk assessment conclusion caused by instantaneous numerical disturbances, and further enhances the comprehensive ability to identify and warn of the cutting machine's operating hazards in advance.
[0145] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.
[0146] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.
[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims
1. A cutting machine operation risk early warning method for cutting a cable protection pipe, characterized in that, The method includes: P1. Acquire the drive motor current signal, feed speed signal, pipe wall deformation feedback signal and spindle vibration signal of the target cutting machine, and perform component analysis on the drive motor current signal to obtain the reference load of the target cutting machine. P2. Based on the feed speed signal and the spindle vibration signal, the reference load is calibrated according to the working condition to obtain the corrected load of the target cutting machine; P3. Sensing the difference trend between the corrected load and the reference load to obtain the load deviation trend of the target cutting machine; P4. Based on the pipe wall deformation feedback signal, perform deformation load collaborative analysis on the load deviation trend to obtain the comprehensive risk deviation of the target cutting machine; P5. Based on the load deviation trend, perform deviation inversion on the benchmark load to obtain the risk judgment range of the target cutting machine; P6. Based on the risk assessment interval, perform risk situation prediction on the comprehensive risk deviation to obtain the risk level of the target cutting machine.
2. The cutting machine operation risk early warning method for cutting a cable protection pipe according to claim 1, characterized in that, The acquisition of the drive motor current signal, feed speed signal, pipe wall deformation feedback signal, and spindle vibration signal of the target cutting machine includes: Collect raw sensor data from the target cutting machine; The original sensor data is filtered to remove interference, resulting in the denoised sampling data of the target cutting machine. The denoised sampling data is sorted to obtain the drive motor current data, feed speed data, pipe wall deformation feedback data and spindle vibration data of the target cutting machine. The drive motor current data, feed speed data, pipe wall deformation feedback data, and spindle vibration data are subjected to amplitude mapping to obtain the drive motor current signal, feed speed signal, pipe wall deformation feedback signal, and spindle vibration signal of the target cutting machine.
3. The cutting machine operation risk early warning method for cutting a cable protection pipe according to claim 1, characterized in that, The step of performing component analysis on the drive motor current signal to obtain the reference load of the target cutting machine includes: The drive motor current signal is decomposed hierarchically to obtain the dimensional component data of the target cutting machine; The effective load of the target cutting machine is obtained by performing steady-state screening on the scale component data; Based on the effective load, the amplitude of the scale component data is merged to obtain the reference load of the target cutting machine.
4. The cutting machine operation risk early warning method for cutting a cable protection pipe according to claim 1, characterized in that, The step of performing condition-related calibration on the reference load based on the feed speed signal and the spindle vibration signal to obtain the corrected load of the target cutting machine includes: The feed speed signal and the spindle vibration signal are identified to be in the same direction to obtain the directional scale of the target cutting machine; Based on the directional scale, the difference between the feed speed signal and the spindle vibration signal is measured to obtain the deviation criterion of the target cutting machine; Based on the directional scale and the deviation criterion, the feed speed signal and the spindle vibration signal are jointly determined to obtain the operating attributes of the target cutting machine; Based on the operating attributes, the baseline load is corrected to obtain the corrected load for the target cutting machine.
5. The cutting machine operation risk early warning method for cutting a cable protection pipe according to claim 1, characterized in that, The step of sensing the difference trend between the corrected load and the reference load to obtain the load deviation trend of the target cutting machine includes: The deviation component of the target cutting machine is obtained by decomposing the difference between the modified load and the reference load. The deviation component is morphologically segmented to obtain the variation segments of the deviation component; By performing correlation evolution analysis on the changed segments, the load deviation trend of the target cutting machine is obtained.
6. The cutting machine operation risk early warning method for cutting a cable protection pipe according to claim 5, characterized in that, The correlation evolution analysis of the changed segments to obtain the load deviation trend of the target cutting machine includes: Feature extraction is performed on the changed segment to obtain the morphological feature information of the changed segment; Based on the morphological feature information, similarity clustering is performed on the changed segments to obtain the classification representation of the changed segments; Based on the classification representation, the temporal sequence of the changed segments is deduced to obtain the connection relationship of the changed segments; Based on the aforementioned connection relationship, the trajectory of the changing segment is reconstructed to obtain the load deviation trend of the target cutting machine.
7. The cutting machine operation risk early warning method for cutting a cable protection pipe according to claim 1, characterized in that, The method of performing deformation-load coordinated analysis on the load deviation trend based on the pipe wall deformation feedback signal to obtain the comprehensive risk deviation of the target cutting machine includes: The load amplitude of the target cutting machine is obtained by parametric analysis of the load deviation trend. Numerical analysis is performed on the pipe wall deformation feedback signal to obtain the deformation amplitude of the target cutting machine; The load amplitude and the deformation amplitude are coupled and measured to obtain the comprehensive risk deviation of the target cutting machine. The formula for calculating the comprehensive risk deviation is as follows: ; wherein, represents the comprehensive risk deviation degree, represents the load amplitude, represents the deformation amplitude, represents a preset deformation matching coefficient, represents the maximum deviation amplitude of the load deviation trend in the entire time domain, represents a natural exponential function.
8. The cutting machine operation risk early warning method for cutting a cable protection pipe according to claim 1, characterized in that, The step of performing deviation inversion on the benchmark load based on the load deviation trend to obtain the risk assessment range of the target cutting machine includes: Deviation sampling is performed on the load deviation trend to obtain the deviation distribution of the target cutting machine; Based on the reference load, a reference offset mapping is performed on the deviation distribution to obtain the offset range of the target cutting machine; The risk assessment range of the target cutting machine is obtained by calibrating the boundary of the offset range.
9. The cutting machine operation risk early warning method for cutting a cable protection pipe according to claim 8, characterized in that, The step of performing a reference offset mapping on the deviation distribution based on the reference load to obtain the offset range of the target cutting machine includes: The deviation values in the deviation distribution are time-series anchored to obtain the time-series position of the deviation of the target cutting machine; Based on the deviation time sequence position, the amplitude of the reference load is sampled to obtain the deviation amplitude sequence of the reference load; Extreme value picking is performed on the deviation amplitude sequence to obtain the maximum and minimum deviation amplitudes of the target cutting machine; Based on the maximum deviation amplitude and the minimum deviation amplitude, the boundary of the reference load is defined to obtain the offset range of the target cutting machine.
10. The method of claim 1, wherein the method further comprises: determining a risk of the cutting machine operating based on the at least one of the cutting machine operating parameter and the cutting machine operating condition. The step of predicting the risk situation based on the comprehensive risk deviation degree based on the risk assessment interval to obtain the risk level of the target cutting machine includes: Based on the risk assessment interval, the location of the comprehensive risk deviation is identified to obtain the interval relative position of the comprehensive risk deviation; Based on the relative position within the interval, the trend of the comprehensive risk deviation is extrapolated to obtain the risk trend of the target cutting machine; The risk trends are summarized and integrated to obtain the risk situation of the target cutting machine; Based on a preset risk level library, the risk situation is hierarchically matched to obtain a risk level of the target cutting machine.
