Pole skeleton wire winding control method and system based on real-time tension feedback

Abstract

This invention discloses a method and system for controlling the wire winding of a power pole frame based on real-time tension feedback. Specifically, it relates to the field of wire winding tension control technology in the manufacturing of prestressed concrete power poles, and is used to solve the problems of tension feedback distortion and uneven prestressing caused by the dynamic changes in the friction coefficient between the steel wire and the frame in existing technologies. By acquiring the steel wire tension feedback value and the axial position information of the winding head in real time, the system judges control anomalies based on tension deviations. When an anomaly occurs, it continuously calculates the higher-order statistics of the tension value and identifies the directional change of the friction state based on its evolution trajectory. It correlates this evolution feature with the axial position to determine the spatial propagation characteristics of the friction state change, evaluates the degree of tension transmission distortion based on this feature, and finally dynamically compensates and adjusts the tension control system parameters according to the degree of distortion. This method can effectively identify and compensate for tension distortion caused by time-varying friction, and improve the uniformity and accuracy of prestressing application.

Description

Technical Field

[0001] This invention relates to the field of wire winding tension control technology in the manufacture of prestressed concrete poles, and more specifically, to a method and system for controlling the wire winding of the pole skeleton based on real-time tension feedback. Background Technology

[0002] In the manufacturing process of prestressed concrete poles, to improve the crack resistance and load-bearing capacity of the components, a common technique is to spirally wind prestressed steel wires onto the surface of the reinforcing steel skeleton. The core of this process lies in applying and maintaining a constant design tension on the steel wires during winding to ensure that it is uniformly converted into effective prestress after the pole is formed. In existing technologies, tension sensors are commonly installed at the wire unwinding or tensioning mechanism of the winding machine to monitor the steel wire tension in real time, and the tensioning device is adjusted based on this feedback signal to maintain stable tension.

[0003] However, in actual winding, the frictional resistance at the interface between the steel wire and the reinforcing cage is not constant; it dynamically changes due to the surface condition of the cage, environmental conditions, and the continuous winding process itself. This time-varying frictional characteristic causes fluctuations in the tension transfer efficiency from the tension detection point to the actual contact point between the steel wire and the cage. This results in a lag and deviation between the detected tension value and the effective tension actually acting on the cage. The control system adjusts based on the distorted feedback signal, making it difficult to ensure that the prestress is accurately applied and distributed according to design requirements throughout the entire length of the pole. This directly affects the consistency of the mechanical properties and long-term durability of the final product. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a method and system for controlling the wire winding of a pole skeleton based on real-time tension feedback to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] The method for controlling the wire winding of a pole skeleton based on real-time tension feedback includes the following steps:

[0007] S1. During the wire winding process, the wire tension feedback value monitored by the tension sensor is obtained in real time, and the axial position information of the wire winding head on the pole frame is obtained simultaneously.

[0008] S2. Based on the deviation between the wire tension feedback value and the preset tension setting value, determine whether there is an abnormality in tension control;

[0009] S3. When an abnormality in tension control is detected, the higher-order statistics of the wire tension feedback value are continuously calculated within the sliding time window, and the direction of the friction state between the wire and the pole frame is identified based on the evolution trajectory of the higher-order statistics.

[0010] S4. When a change in direction occurs, the evolution characteristics of higher-order statistics are correlated with the axial position information of the winding head to determine the spatial propagation characteristics of the change in friction state.

[0011] S5. Based on the spatial propagation characteristics of changes in friction state, assess the degree of tension transmission distortion caused by time-varying friction coefficient;

[0012] S6. Based on the degree of tension transmission distortion, dynamically compensate and adjust the parameters of the tension control system.

[0013] Furthermore, S1 includes:

[0014] The wire tension feedback value monitored by the tension sensor is filtered in real time to obtain the wire tension feedback value used for control.

[0015] The axial position information of the winding head is calibrated in real time to obtain the axial position information used for control.

[0016] Furthermore, S2 includes:

[0017] Calculate the real-time deviation between the wire tension feedback value used for control and the preset tension setting value;

[0018] Calculate the moving range of this real-time deviation over a continuous time interval;

[0019] The moving range is compared with the dynamic control limits established based on historical stable winding process data;

[0020] When the range of motion continuously exceeds the dynamic control limit, it is judged that an abnormality in tension control has occurred.

[0021] Furthermore, the dynamic control limit is dynamically generated based on the moving range statistical characteristics of the difference between the wire tension feedback value used for control and its corresponding preset tension setting value in historical stable winding process data.

[0022] Furthermore, S3 includes:

[0023] After determining that an abnormality in tension control has occurred, the skewness and kurtosis of the wire tension feedback value used for control are continuously calculated within the sliding time window.

[0024] Using skewness and kurtosis as coordinates, an evolution trajectory within a sliding time window is constructed on a two-dimensional plane formed by the skewness axis and the kurtosis axis.

[0025] Trajectory clustering analysis is performed on the points on the two-dimensional plane of the evolution trajectory to extract the movement direction and clustering pattern of the evolution trajectory;

[0026] When the direction of movement of the evolution trajectory shows a trend of continuously moving away from the pre-defined stable friction state cluster center region, it is identified as a directional change in the friction state.

[0027] Furthermore, S4 includes:

[0028] During the period when the friction state is identified as a directional change, the movement direction of the evolution trajectory and the time point corresponding to the aggregation pattern are recorded simultaneously, and the axial position sequence formed during this period is extracted for control purposes.

[0029] Analyze the correlation between the rate of change of the direction of movement of the evolution trajectory and the changes in the axial position sequence;

[0030] Based on the correlation, it is determined whether the change in friction state is a gradual propagation mode along the pole frame axis or a local concentrated propagation mode, which serves as a spatial propagation characteristic.

[0031] Furthermore, the correlation between the rate of change of the direction of movement of the evolution trajectory and the change of the axial position sequence is analyzed by calculating the Pearson correlation coefficient between the sequence of changes in the movement angle of the evolution trajectory on a two-dimensional plane composed of the skewness axis and the kurtosis axis and the amount of change in the axial position sequence.

[0032] Furthermore, S5 includes:

[0033] Based on the determined gradual propagation mode or localized centralized propagation mode, and combined with the movement direction and aggregation mode of the evolution trajectory, a model for assessing the degree of tension transmission distortion is established.

[0034] Based on the axial position range corresponding to the spatial propagation characteristics and the degree of deviation of the evolution trajectory from the stable state, a comprehensive quantitative value of the degree of tension transmission distortion is calculated through the tension transmission distortion degree evaluation model.

[0035] Furthermore, S6 includes:

[0036] Based on the magnitude of the quantified value of the tension transmission distortion and the spatial propagation characteristics, a corresponding dynamic compensation adjustment amount is generated;

[0037] When the spatial propagation characteristic is a progressive propagation mode, the feedforward compensation parameters of the tension control system are corrected by dynamic compensation adjustment.

[0038] When the spatial propagation characteristic is a local centralized propagation mode, the feedback control gain of the tension control system is adjusted by the dynamic compensation adjustment amount.

[0039] On the other hand, the present invention provides a pole frame winding control system based on real-time tension feedback, comprising the following modules:

[0040] The data acquisition module is used to acquire the wire tension feedback value monitored by the tension sensor in real time during the wire winding process, and simultaneously acquire the axial position information of the wire winding head on the pole frame.

[0041] The anomaly detection module is used to determine whether an anomaly in tension control has occurred based on the deviation between the wire tension feedback value and the preset tension setting value.

[0042] The state recognition module is used to continuously calculate the higher-order statistics of the wire tension feedback value within the sliding time window when an abnormality in tension control is detected, and to identify whether the friction state between the wire and the pole frame has changed direction based on the evolution trajectory of the higher-order statistics.

[0043] The feature association module is used to associate the evolution characteristics of higher-order statistics with the axial position information of the winding head when a change in direction occurs, so as to determine the spatial propagation characteristics of the change in friction state.

[0044] The distortion assessment module is used to assess the degree of tension transmission distortion caused by the time-varying friction coefficient, based on the spatial propagation characteristics of changes in friction state.

[0045] The compensation and adjustment module is used to dynamically compensate and adjust the parameters of the tension control system according to the degree of tension transmission distortion.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] 1. By constructing a complete technology chain from anomaly detection, state diagnosis, spatial positioning to quantitative assessment and precise compensation, proactive insight and closed-loop suppression of the deep-seated disturbance of time-varying friction coefficient are achieved. Compared with traditional control that relies solely on instantaneous tension feedback, introducing and analyzing the higher-order statistical characteristics and evolution trajectory of wire tension in the time domain can keenly capture the migration of the probability distribution of tension signals caused by fundamental changes in friction state. This allows the system to distinguish between random disturbances and trend changes indicating a systematic deterioration in the tension transmission relationship, thus providing a reliable decision-making basis for proactive intervention.

[0048] 2. By correlating the evolution characteristics of the friction state with the axial spatial information of the winding process, we can not only determine when the anomaly occurs, but also how it propagates. This allows the final dynamic compensation adjustment to act differently on the feedforward or feedback links of the tension control system based on the spatial propagation mode (gradual or localized). This can significantly improve the uniformity and accuracy of prestress applied throughout the entire length of the pole frame, thereby ensuring the consistency of the mechanical properties and structural durability of the concrete pole after molding, and effectively solving the problem of uncontrolled prestress distribution caused by feedback distortion. Attached Figure Description

[0049] Figure 1 The flowchart shows the pole frame winding control method based on real-time tension feedback according to the present invention.

[0050] Figure 2 This is a schematic diagram of the structure of the pole frame winding control system based on real-time tension feedback according to the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0052] Example 1: Figure 1 The present invention provides a method for controlling the wire winding of a pole skeleton based on real-time tension feedback, which includes the following steps:

[0053] S1. During the wire winding process, the wire tension feedback value monitored by the tension sensor is obtained in real time, and the axial position information of the wire winding head on the pole frame is obtained simultaneously.

[0054] S2. Based on the deviation between the wire tension feedback value and the preset tension setting value, determine whether there is an abnormality in tension control;

[0055] S3. When an abnormality in tension control is detected, the higher-order statistics of the wire tension feedback value are continuously calculated within the sliding time window, and the direction of the friction state between the wire and the pole frame is identified based on the evolution trajectory of the higher-order statistics.

[0056] S4. When a change in direction occurs, the evolution characteristics of higher-order statistics are correlated with the axial position information of the winding head to determine the spatial propagation characteristics of the change in friction state.

[0057] S5. Based on the spatial propagation characteristics of changes in friction state, assess the degree of tension transmission distortion caused by time-varying friction coefficient;

[0058] S6. Based on the degree of tension transmission distortion, dynamically compensate and adjust the parameters of the tension control system.

[0059] S1. During the wire winding process, the wire tension feedback value monitored by the tension sensor is acquired in real time, and the axial position information of the winding head on the pole frame is acquired simultaneously. The specific implementation is as follows:

[0060] During the wire winding process, a tension sensor continuously monitors the wire tension, outputting a continuous wire tension feedback value that includes useful signals and high-frequency measurement noise. To obtain a stable and reliable wire tension feedback value for control, the original monitoring signal needs to be filtered in real time. Specifically, a first-order digital low-pass filter is used to process the wire tension feedback value. The filter's cutoff frequency is determined based on the rotation frequency of the winding machine's spindle and the fundamental frequency of the possible longitudinal vibration of the wire. The aim is to effectively suppress high-frequency noise caused by mechanical vibration and electromagnetic interference, while retaining the low-frequency components of the true tension change. For example, the cutoff frequency can be set to 5 times the spindle rotation frequency, but lower than the fundamental frequency of the wire's longitudinal vibration. In each control cycle, the filter receives the latest tension sensor sample value and, based on the filtered output value of the previous cycle and the set filtering time constant, recursively calculates the filtered tension value for the current cycle. This value is the wire tension feedback value used for control. This processing is continuous, ensuring that the output wire tension feedback value for control is smooth and reflects the tension trend in real time.

[0061] Simultaneously, to accurately determine the current position of the winding head relative to the pole frame, the axial position information of the winding head needs to be acquired synchronously. This information is typically provided by a linear encoder mounted on the feed screw or linear guide of the winding machine. The encoder outputs pulse signals, each pulse corresponding to a fixed axial displacement. However, due to mechanical backlash, thermal deformation, or potential zero-point drift of the encoder itself, the initially read axial position information may contain accumulated errors. Therefore, real-time calibration is required to obtain accurate axial position information for control. The calibration process relies on a preset, physically defined axial position reference point on the pole frame, such as a specific mark at the end of the frame. Whenever the winding head passes this axial position reference point under the drive of the control system, regardless of the encoder reading at that time, the system forcibly resets the axial position information used for control to the known coordinate value of that reference point, for example, setting the axial position coordinates to zero. Within the travel distance between the two reference points, the axial position information used for control is updated in real time by accumulating the equivalent number of pulses emitted by the encoder since the last calibration. By performing this calibration operation periodically or on a triggered basis, the accumulated error of the axial position information can be eliminated, ensuring that it always maintains a precise correspondence with the actual physical position of the winding head on the pole frame, thereby obtaining reliable axial position information for control.

[0062] S2. Based on the deviation between the wire tension feedback value and the preset tension setting value, determine whether an abnormality in tension control has occurred. The specific implementation is as follows:

[0063] After obtaining the wire tension feedback value used for control, the real-time deviation is calculated by subtracting the preset tension setpoint from the feedback value. The preset tension setpoint is a target wire tension value, set in advance according to the pole design requirements, intended to remain constant during winding, and its unit is Newtons. The real-time deviation is calculated synchronously in each control cycle, and the real-time deviation equals the wire tension feedback value used for control minus the preset tension setpoint; the unit of the real-time deviation is Newtons.

[0064] To identify abnormal patterns from fluctuations in real-time deviation, it is necessary to calculate the moving range of the real-time deviation over a continuous time interval. The continuous time interval refers to a sliding time window of fixed length. The length of the sliding time window needs to be set according to the process speed of the winding machine to ensure that it contains enough data points to reflect the fluctuation characteristics, while avoiding an excessively long window that would lead to sluggish response. For example, the process speed of the winding machine is typically a certain number of turns per minute, and the corresponding axial feed rate is known. The length of the sliding time window can be set to cover at least 1 to 3 times the time required to complete one full turn. Specifically, the length of the sliding time window can be set to a value between 3 and 10 seconds, such as 5 seconds. The moving range is calculated as follows: at each calculation moment, all real-time deviation data points are extracted from the current moment back to the length of the sliding time window. The maximum and minimum values ​​among these real-time deviation data points are found, and then the maximum value is subtracted from the minimum value. The difference is the moving range corresponding to the current moment. The unit of the moving range is the same as that of the real-time deviation, namely Newtons. As time progresses and the sliding time window moves, a continuous moving range sequence is obtained, which reflects the fluctuation range of the real-time deviation over a short period of time.

[0065] To determine whether the moving range is abnormal, it is necessary to compare the calculated moving range with the dynamic control limits. The dynamic control limits are established based on historical stable winding process data. Historical stable winding process data refers to the recorded wire tension feedback values ​​used for control and their corresponding preset tension settings during historical production stages where equipment was properly debugged, process parameters were stable, and the produced poles were of acceptable quality. The selection of historical stable winding process data should cover multiple production batches and different environmental conditions to ensure the universality of statistical characteristics. From the historical stable winding process data, a sequence of differences between the wire tension feedback values ​​used for control and their corresponding preset tension settings can be calculated. Then, according to the sliding time window length and calculation period of the real-time production intermediate sample, the moving range sequence of historical differences can be calculated. Finally, statistical characteristic analysis is performed on the historical moving range sequence, such as calculating the mean and standard deviation of the historical moving range sequence. The dynamic control limits are generated based on these statistical characteristics. For example, the dynamic control limits can be set as the mean of the historical moving range plus twice the standard deviation of the historical moving range, or as the mean of the historical moving range plus three times the standard deviation of the historical moving range. The choice of factor depends on the required balance between control sensitivity and false alarm rate; for example, two factors can be used when higher sensitivity is required, and three factors can be used when a lower false alarm rate is required. The dynamic control limit is a numerical boundary, measured in Newtons, and characterizes the upper limit range of normal fluctuations in the moving range under stable production conditions.

[0066] The logic for determining whether a tension control anomaly has occurred lies in monitoring whether the moving range consistently exceeds the dynamic control limit. A consistent exceedance means that the moving range is greater than the dynamic control limit for multiple consecutive calculation cycles. For example, a sustained judgment threshold can be set; this threshold is an integer representing the number of calculation cycles in which the dynamic control limit is consistently exceeded. The setting of this threshold needs to consider the response speed of the control system and the reliability of anomaly confirmation. For instance, the threshold can be set to a value between 3 and 10, such as 5. In each control cycle, the system compares the currently calculated moving range with the dynamic control limit. If the moving range is greater than the dynamic control limit, an exceedance event is recorded. The system maintains a counter, initially set to zero. The counter increments by 1 when an exceedance event occurs and resets to zero when the moving range is lower than or equal to the dynamic control limit. If the counter reaches or exceeds the sustained judgment threshold, for example, a counter reaching 5, a tension control anomaly is determined to have occurred. This method effectively distinguishes between short-lived random disturbances and persistent abnormal fluctuations that characterize a fundamental change in the friction state, because short-lived disturbances typically do not cause the moving range to consistently exceed the dynamic control limit established based on long-term stable historical data. The determination of tension control abnormalities provides the triggering conditions for subsequent steps.

[0067] S3. When an abnormality in tension control is detected, the higher-order statistics of the wire tension feedback value are continuously calculated within the sliding time window, and the directionality of the friction state between the wire and the pole frame is identified based on the evolution trajectory of the higher-order statistics. Specifically, the implementation is as follows:

[0068] Upon detecting an anomaly in tension control, the system initiates a deep analysis of the control wire tension feedback value to detect fundamental changes in the friction state between the wire and the pole frame. The core of this analysis lies in calculating the higher-order statistics of the control wire tension feedback value and observing its dynamic evolution. The skewness and kurtosis of the control wire tension feedback value are continuously calculated within a sliding time window. This sliding time window acts as a continuously advancing data buffer. The length of the sliding time window must balance the ability to capture signal change trends with real-time calculation. The length of the sliding time window depends on the winding machine's process speed and the control system's sampling frequency to ensure that the window contains a sufficient number of data points to reflect statistical characteristics, while avoiding excessively long windows that could cause response delays. For example, if the winding machine's control system sampling frequency is 100 Hz and the process speed makes winding one turn take approximately 2 seconds, then the length of the sliding time window can be set to cover the time of 1 to 2 turns of winding, such as a value between 2 and 5 seconds. Skewness is a statistic used to measure the asymmetry of data distribution. It is calculated as follows: for a series of wire tension feedback values ​​used for control within a sliding time window, first calculate the mean and standard deviation of the series; then calculate the average of the cube of the difference between each data point and the mean; finally, divide this average by the cube of the standard deviation. Kurtosis is a statistic used to measure the steepness or flatness of data distribution. It is calculated as follows: for a series of wire tension feedback values ​​used for control within the same sliding time window, calculate the average of the fourth power of the difference between each data point and the mean; then divide this average by the fourth power of the standard deviation, and subtract 3. Both skewness and kurtosis are dimensionless pure numbers. The calculation process is performed continuously as the sliding time window moves, resulting in a pair of time-varying skewness and kurtosis sequences.

[0069] Using the calculated skewness value at each moment as the x-axis and the kurtosis value as the y-axis, each pair of skewness and kurtosis data points is plotted on a two-dimensional plane formed by the skewness and kurtosis axes. As time progresses, these continuous data points form a path on this two-dimensional plane, which is the evolution trajectory within the sliding time window. The evolution trajectory visually demonstrates how the probability distribution of the control wire tension feedback value dynamically evolves over time.

[0070] The purpose of trajectory clustering analysis on the points traced by the evolutionary trajectory in a two-dimensional plane is to extract the macroscopic behavioral patterns of the trajectory. Trajectory clustering analysis does not cluster individual data points, but rather classifies trajectory segments or the overall morphological characteristics of the trajectory. One implementation method is to divide the evolutionary trajectory into continuous, overlapping short line segments, each consisting of several consecutive data points. The length of each short line segment can be set to contain 10 to 20 consecutive data points. A feature vector is calculated for each short line segment, which may include the segment's orientation angle, average curvature, and the skewness and kurtosis values ​​of its starting and ending points. Then, a clustering algorithm, such as K-means clustering, is used to cluster these feature vectors. Through clustering, the main directions of movement of the evolutionary trajectory in the two-dimensional plane can be identified, such as directions that predominantly tend towards increasing skewness or decreasing kurtosis; these are referred to as the movement directions of the evolutionary trajectory. Meanwhile, the clustering results also reveal the clustering patterns of data points on a two-dimensional plane, such as whether the data points are densely clustered in a small area, scattered in a strip-shaped area, or forming multiple separate clusters.

[0071] To determine whether the friction state has undergone a directional change, the currently observed evolutionary trajectory behavior needs to be compared with a baseline. This baseline is a pre-defined cluster center region for stable friction states. The stable friction state cluster center region is established based on a large amount of historical stable production data. The method for establishing this region is as follows: collect wire tension feedback values ​​used for control when the equipment is operating normally, the process is stable, and the product quality is qualified; calculate the skewness and kurtosis using the same method; plot these historical skewness and kurtosis data points on the same two-dimensional plane; then perform spatial cluster analysis on these historical data points to identify the most prominent clustering areas. The center or core area of ​​this region is then designated as the stable friction state cluster center region. For example, this region can be a circular area with a specific skewness and kurtosis value as its center and a certain statistical distance as its radius. This radius can be set to twice the standard deviation of the historical data point distribution.

[0072] The logic for identifying directional changes in friction states is as follows: observe whether the current trajectory's movement direction shows a trend of continuously moving away from the pre-defined stable friction state cluster center region. This trend of continuous movement needs to be determined using quantitative indicators. First, calculate the Euclidean distance between each data point on the evolution trajectory and the center point of the stable friction state cluster center region, obtaining a distance sequence. Then, analyze whether this distance sequence shows a monotonically increasing trend over multiple consecutive calculation cycles. The threshold used to determine the persistence of the trend is called the trend persistence length threshold, which is an integer representing the number of consecutive monotonically increasing calculation cycles. The setting of the trend persistence length threshold needs to consider the response speed of the control system and the reliability of anomaly confirmation. For example, the trend persistence length threshold can be set to a value between 5 and 12 consecutive calculation cycles, such as 8. Simultaneously, calculate the angle of the evolution trajectory's movement direction, defined as the angle between the vector pointing from the center point of the stable friction state cluster center region to the current trajectory data point and the positive direction of the skewness axis. If the angle changes very little over several consecutive periods, for example, the angle change range is less than 30 degrees, and it points in a direction outside the cluster center of the stable friction state, then the direction of movement is confirmed to be directional and moving away. When the distance sequence monotonically increases for more than a certain number of consecutive periods (e.g., 8 consecutive periods), and the angle of movement consistently points outward, then a directional change in the friction state is identified. This directional change means that the contact friction characteristics between the steel wire and the pole frame are undergoing a systematic and trend-driven deterioration or transformation, rather than random fluctuations. The identification of a directional change in the friction state provides a basis for subsequent analysis of its spatial propagation characteristics.

[0073] S4. When a change in direction occurs, the evolution characteristics of higher-order statistics are correlated with the axial position information of the winding head to determine the spatial propagation characteristics of the change in friction state. Specifically, this is implemented as follows:

[0074] Once a directional change in friction state is detected, the system enters the spatial localization analysis phase for this change. During the duration of this directional change, the system synchronously records the time points corresponding to the movement direction and clustering pattern of the evolution trajectory, and extracts the axial position sequence formed within this time period for control purposes. The duration of the directional change refers to the time interval from when the directional change determination condition is met until it is no longer met. The movement direction of the evolution trajectory has been extracted in previous steps through trajectory clustering analysis; for example, the movement direction can be quantified as an angle value, representing the average direction of the trajectory on a two-dimensional plane. The clustering pattern describes the distribution characteristics of data points on a two-dimensional plane; for example, the clustering pattern can be quantified as the cluster center coordinates or the dispersion of the clusters. Time points refer to the movement direction value and clustering pattern value corresponding to each calculation cycle; the system records these values ​​along with their corresponding timestamps. Simultaneously, the system extracts the axial position information for control within the same time interval. This axial position information comes from a real-time calibrated axial position sensor, such as a linear encoder. The axial position information used for control in each calculation cycle is arranged in chronological order to form an axial position sequence. The unit of the axial position sequence can be millimeters, representing the axial coordinates of the wire winding head relative to a reference point on the pole frame. Through synchronous recording, it is ensured that the movement direction and aggregation pattern of each time point are strictly aligned with the position values ​​in the axial position sequence in time, providing a consistent data foundation for subsequent correlation analysis.

[0075] This study analyzes the correlation between the rate of change of the movement direction of the evolution trajectory and the changes in the axial position sequence. The rate of change of the movement direction refers to how quickly the movement direction angle changes over time, specifically characterized by calculating the movement angle change sequence. The calculation method for the movement angle change sequence is as follows: First, extract the movement direction angle sequence from the recorded time points, i.e., the movement direction angle value corresponding to each calculation cycle. Then, calculate the difference in movement direction angles between adjacent calculation cycles to obtain the change amount sequence. To smooth random fluctuations, a moving average filter can be applied to the change amount sequence, for example, using a moving average with a window length of 3 to 5 cycles. The change in the axial position sequence refers to the rate of change of the axial position over time, i.e., the axial movement speed of the winding head. The calculation method for the change in the axial position sequence is as follows: calculate the difference in axial position values ​​between adjacent calculation cycles to obtain the change amount sequence of the axial position. This change amount sequence directly reflects the instantaneous axial movement speed of the winding head, and the unit can be millimeters per second. To analyze the correlation between the rate of change of the movement direction and the changes in the axial position sequence, it is necessary to calculate the Pearson correlation coefficient between the movement angle change sequence and the change amount of the axial position sequence. The calculation of the Pearson correlation coefficient requires two sequences of equal length. The sequences of changes in the movement angle and the axial position are time-aligned. The specific calculation process for the Pearson correlation coefficient is as follows: First, calculate the average of the movement angle change sequence and the average of the axial position change sequence. Then, calculate the deviation of each sequence element from its average. Next, calculate the average of the products of the two sequence deviations, and simultaneously calculate the square root of the average of the squares of the two sequence deviations. Finally, divide the average of the deviation products by the product of the two square roots to obtain the Pearson correlation coefficient value. The Pearson correlation coefficient ranges from -1 to +1, and its absolute value indicates the strength of the linear correlation; positive and negative indicate the direction of the correlation. To determine whether the correlation is significant, a correlation threshold needs to be set. The correlation threshold is a value between 0 and 1, for example, it can be set to 0.6 or 0.7. If the absolute value of the calculated Pearson correlation coefficient is greater than the correlation threshold, a significant correlation is considered to exist between the rate of change of the movement direction of the evolution trajectory and the change in the axial position sequence; otherwise, the correlation is considered insignificant. The correlation threshold is set based on statistical significance tests, such as a critical value obtained by looking up the sequence length in a table at a 95% confidence level, or an empirical value determined through experiments using historical data.

[0076] Based on the calculated Pearson correlation coefficient and its significance, the spatial propagation characteristics of friction state changes were determined. Spatial propagation characteristics describe the expansion of friction state changes along the pole frame axis, and are divided into a gradual propagation mode and a locally concentrated propagation mode. The gradual propagation mode refers to the gradual development of friction state changes as the winding head moves axially, meaning the change in movement direction is synchronized with the change in axial position. The locally concentrated propagation mode refers to the friction state change occurring and persisting near a fixed axial position, without significantly spreading with the movement of the winding head. The criteria for determining the mode include the value and sign of the Pearson correlation coefficient, as well as the range of change in the axial position sequence. If the absolute value of the Pearson correlation coefficient is greater than the correlation threshold and the sign is positive, it indicates that the rate of change in movement direction is positively correlated with the axial movement speed; that is, the faster the winding head moves, the faster the direction of friction state change, which is consistent with the characteristics of a gradual propagation mode. Simultaneously, the range of change in the axial position sequence during the observation period is examined. If the range is large, for example, exceeding 10% of the pole frame length, it further supports the gradual propagation mode. If the absolute value of the Pearson correlation coefficient is less than the correlation threshold, it indicates that the rate of change in the direction of movement has no significant linear relationship with the axial movement speed. Simultaneously, the range of change in the axial position sequence is examined. If the range is very small, for example, less than 1% of the pole frame length, it indicates that the change in friction state is confined to a narrow axial region, consistent with a localized concentrated propagation pattern. Furthermore, the change in the aggregation pattern can be considered. For a gradual propagation pattern, the aggregation pattern may drift slowly with the axial position; for a localized concentrated propagation pattern, the aggregation pattern may remain relatively stable. By comprehensively considering the Pearson correlation coefficient, the range of change in the axial position sequence, and the aggregation pattern behavior, the system determines whether the current change in friction state belongs to a gradual or localized concentrated propagation pattern, and outputs this determination as a spatial propagation characteristic. The determination of the spatial propagation characteristic provides crucial spatial dimensional information for subsequent assessment of the degree of tension transmission distortion.

[0077] S5. Based on the spatial propagation characteristics of changes in friction state, assess the degree of tension transmission distortion caused by the time-varying friction coefficient. Specifically, this is implemented as follows:

[0078] Based on the determined gradual or locally concentrated propagation pattern, and combining the movement direction and clustering pattern of the evolution trajectory, a tension transmission distortion assessment model is established. This model is a mathematical relationship used to transform the qualitative characteristics and spatial information of friction state changes into a numerical value characterizing the severity of distortion. The model's establishment depends on the classification of spatial propagation characteristics. For the gradual propagation pattern, friction state changes gradually develop with the axial movement of the winding head; therefore, the tension transmission distortion assessment model needs to reflect the cumulative effect of distortion along the axial direction. For the locally concentrated propagation pattern, friction state changes are concentrated near a certain axial position; therefore, the model needs to reflect the distortion intensity in that local area. The movement direction of the evolution trajectory has already been extracted in step S3; for example, the movement direction can be represented as an angle value, which may continuously change during the directional change of friction state. The clustering pattern of the evolution trajectory has also been extracted in step S3. The clustering pattern can be quantified as the radius of the clusters or the standard deviation of the data point distribution, used to represent the dispersion of skewness and kurtosis data points. When establishing a model to assess the degree of tension transmission distortion, it is necessary to assign weights to the contributions of the movement direction and the aggregation pattern. The weights are set based on historical data analysis or experimental calibration. For example, multiple sets of historical production data can be collected, including known cases of poles with uneven prestress due to friction issues, as well as normal pole cases. For each set of historical data, the corresponding spatial propagation characteristics, the movement direction sequence of the evolution trajectory, and the aggregation pattern sequence are calculated according to steps S1 to S4. Then, through expert evaluation or back-calculation based on subsequent pole mechanical performance test results, a reference level or value for the degree of tension transmission distortion is assigned to each historical case. Using this historical data, a mapping relationship is established from movement direction features, aggregation pattern features, and spatial propagation feature types to the distortion reference value through multiple linear regression analysis or machine learning methods such as support vector regression. This mapping relationship constitutes the tension transmission distortion assessment model. The features in the model can be extracted from the movement direction sequence and the aggregation pattern sequence; for example, the movement direction feature could be the average rate of change of the movement direction angle during the observation period, and the aggregation pattern feature could be the maximum value of the aggregation pattern during the observation period. The model can take the form of a linear combination. For example, for a progressive propagation model, the base value of the distortion is equal to the movement direction feature multiplied by weight W1, plus the clustering pattern feature multiplied by weight W2; for a locally concentrated propagation model, the base value of the distortion is equal to the movement direction feature multiplied by weight W3, plus the clustering pattern feature multiplied by weight W4. Weights W1, W2, W3, and W4 are obtained through regression of historical data, for example, by using least squares fitting to minimize the error between the model output and the reference value. The model may also include a constant term.In addition, the model needs to adapt to different axial position ranges, so the final output of the model is the product or weighted sum of the base value of the distortion and an axial position range factor.

[0079] Based on the axial position range corresponding to the spatial propagation characteristics and the degree of deviation of the evolution trajectory from the stable state, a comprehensive quantitative value of the degree of tension transmission distortion is calculated through a tension transmission distortion assessment model. The axial position range corresponding to the spatial propagation characteristics refers to the axial position interval traversed by the winding head during the duration of the directional change in the friction state. From the axial position sequence, the minimum and maximum axial position values ​​within that time period are found, and the axial position range is obtained by subtracting the minimum axial position value from the maximum axial position value. The unit of the axial position range is millimeters. The axial position range reflects the spatial span of the impact of the friction state change. For a gradual propagation mode, the axial position range is usually larger; for a locally concentrated propagation mode, the axial position range is usually smaller. The degree of deviation of the evolution trajectory from the stable state is used to quantify the magnitude of the friction state change in the statistical feature space. The degree of deviation of the evolution trajectory from the stable state is characterized by calculating the average distance between the data point of the current evolution trajectory and the center point of the pre-calibrated stable friction state cluster center region. The specific calculation method is as follows: During the duration of the directional change in the friction state, for each calculation cycle, the Euclidean distance between the skewness and kurtosis data points obtained and the center point of the cluster center region of the stable friction state is calculated. Then, the average of the distances calculated in all cycles is used to obtain the degree of deviation of the evolution trajectory from the stable state. This degree is a dimensionless value; the larger the value, the farther the deviation. To integrate the axial position range and the degree of deviation into the model, it is necessary to define an axial position range factor and a deviation degree factor. The axial position range factor is a function of the axial position range. For example, the axial position range factor can be equal to the axial position range divided by a reference length, which can be the total length of the pole frame or a typical value such as 1000 mm. The deviation degree factor can directly use the calculated value of the degree of deviation of the evolution trajectory from the stable state. Then, the axial position range factor and the deviation degree factor are used as inputs, combined with the previously established tension transmission distortion evaluation model, to calculate the quantified value. For example, for the progressive propagation mode, the quantified value of the overall tension transmission distortion degree is equal to the base value of the distortion degree plus the deviation degree factor multiplied by coefficient K1 plus the axial position range factor multiplied by coefficient K2. For the localized centralized propagation mode, the quantification value equals the base value of distortion plus the deviation factor multiplied by coefficient K3, plus the reciprocal of the axial position range factor multiplied by coefficient K4. This is because in localized centralized propagation, a smaller range may indicate a more concentrated problem and a stronger distortion effect. Coefficients K1, K2, K3, and K4 are also determined through historical data regression analysis to ensure the final quantification value best matches the reference distortion levels of historical cases. During the calculation process, all parameters and intermediate variables are operated using double-precision floating-point numbers to ensure accuracy. The final quantification value of the overall tension transmission distortion is a dimensionless value, typically between 0 and 100. A larger value indicates a more severe tension transmission distortion caused by the time-varying friction coefficient.This quantified value serves as the direct basis for subsequent dynamic compensation adjustments. The entire evaluation process is executed periodically in a dedicated calculation thread within the control system to ensure real-time performance.

[0080] S6. Based on the degree of tension transmission distortion, dynamically compensate and adjust the parameters of the tension control system. Specifically, this is implemented as follows:

[0081] Based on the magnitude of the quantized value of tension transmission distortion and the spatial propagation characteristics, a corresponding dynamic compensation adjustment is generated. The quantized value of tension transmission distortion is a dimensionless value calculated in step S5, typically between 0 and 100. The spatial propagation characteristics have been determined in step S4 as either a gradual propagation mode or a localized centralized propagation mode. The dynamic compensation adjustment is an adjustment value used to modify the internal parameters of the tension control system. The specific method for generating the dynamic compensation adjustment is as follows: First, a baseline adjustment is defined, which is proportional to the quantized value of tension transmission distortion. For example, the baseline adjustment can be set as the quantized value of tension transmission distortion multiplied by a proportionality coefficient Kbase. The proportionality coefficient Kbase is an empirical constant used to map the quantized value to the adjustment scale of the actual control parameters. The determination of the proportionality coefficient Kbase is based on historical debugging data, such as analyzing historical cases to determine how much parameter adjustment corresponds to a certain degree of quantized value to effectively compensate for distortion, and obtaining the value of the proportionality coefficient Kbase through linear fitting. Then, the baseline adjustment is corrected according to the spatial propagation characteristics. For the gradual propagation mode, since distortion develops along the axial direction, compensation needs to be predictable and gradual. Therefore, the dynamic compensation adjustment is equal to the baseline adjustment multiplied by a smoothing factor, which is less than 1 (e.g., 0.7) to make the adjustment action gentler and avoid oscillations. For the locally concentrated propagation mode, since distortion is concentrated, rapid and strong local intervention is required. Therefore, the dynamic compensation adjustment is equal to the baseline adjustment multiplied by a strengthening factor, which is greater than 1 (e.g., 1.3) to amplify the adjustment effect. The specific values ​​of the smoothing factor and strengthening factor are determined through simulation or field experiments to find the most effective adjustment intensity for different propagation modes. The calculation of generating the dynamic compensation adjustment is performed once in each control evaluation cycle to ensure the real-time nature of the adjustment.

[0082] When the spatial propagation characteristic is a progressive propagation mode, the feedforward compensation parameters of the tension control system are corrected using dynamic compensation adjustment. Tension control systems typically employ a composite control structure that includes feedback control and feedforward compensation. The role of feedforward compensation is to output control quantities in advance based on known disturbances or system dynamic characteristics to offset anticipated errors. In this invention, the feedforward compensation parameter specifically refers to the parameter used to predict and compensate for the decrease in tension transmission efficiency caused by time-varying friction coefficients. This parameter can be the gain coefficient of the feedforward controller or a feedforward compensation table. The correction method is as follows: First, the current feedforward compensation parameter value is read and stored in the memory of the control system. Then, the dynamic compensation adjustment is calculated with the current feedforward compensation parameter value. A specific calculation is additive correction, where the new feedforward compensation parameter equals the original feedforward compensation parameter plus the dynamic compensation adjustment. Since time-varying friction coefficients usually lead to a decrease in tension transmission efficiency, feedforward compensation needs to be increased to pre-raise the tension setting; therefore, the dynamic compensation adjustment is usually a positive number, and additive correction ensures the enhancement of the feedforward compensation effect. Another approach is multiplicative correction, where the new feedforward compensation parameter equals the original feedforward compensation parameter multiplied by a coefficient equal to 1 plus the ratio of the dynamic compensation adjustment to a normalized base value. Whether additive or multiplicative correction is used depends on the original feedforward compensation parameter structure of the control system, which needs to be determined during system design. The corrected feedforward compensation parameter is immediately updated in the controller's parameter range and applied to the next control cycle and subsequent winding processes. For the progressive propagation mode, this correction takes effect gradually throughout the identified friction state change region. The system applies the corrected feedforward parameter when the winding head enters the affected axial position range, based on axial position information, thereby achieving proactive compensation for the development trend.

[0083] When the spatial propagation characteristic is a localized concentrated propagation mode, the feedback control gain of the tension control system is adjusted using a dynamic compensation adjustment. The feedback control gain determines the response speed and intensity of the tension control system to tension deviations, such as the proportional gain and integral gain. The purpose of the adjustment is to enhance the error correction capability of the control system in localized areas where friction is abnormally concentrated, and to quickly overcome tension distortion caused by localized high friction or jamming. The adjustment method is as follows: First, determine the feedback control gain term that needs adjustment, usually the proportional gain. Then, calculate the gain adjustment amount based on the dynamic compensation adjustment. The gain adjustment amount can be obtained from the dynamic compensation adjustment amount through a mapping function. For example, the gain adjustment amount can be equal to the dynamic compensation adjustment amount multiplied by a gain adjustment coefficient Kgain, which converts the dynamic compensation adjustment amount into a relative change in gain. For example, if the original proportional gain is Gp, the new proportional gain can be set to Gp multiplied by a factor equal to 1 plus the gain adjustment amount divided by 100. The specific operation can be as follows: First, normalize the dynamic compensation adjustment to the range of 0 to 1, then multiply it by a maximum allowable gain adjustment ratio, such as 20%, to obtain the actual gain adjustment ratio. Next, increase or decrease the current value of the feedback control gain by this adjustment ratio. For local centralized propagation modes, it is usually necessary to enhance the feedback effect to overcome local disturbances, thus increasing the feedback control gain. The adjustment process needs to be cautious to avoid excessive gain leading to system instability. Therefore, a gain upper limit threshold can be set, which is the maximum allowable feedback control gain, for example, 1.5 times the original calibration gain. When calculating the new gain value, it is necessary to ensure that it does not exceed the gain upper limit threshold. Gain adjustment is also completed within the control cycle, and the new feedback control gain value takes effect immediately. When the winding head moves to the local axial position region corresponding to the spatial propagation characteristics, the system applies the adjusted higher gain to provide stronger corrective force; when the winding head leaves this region, the feedback control gain can be gradually restored to the original calibration value. The restoration process can use a ramp function to avoid abrupt changes. This dynamic, spatially linked gain adjustment effectively suppresses localized, concentrated friction anomalies. The entire dynamic compensation and adjustment process operates in a closed loop. The tension control system uses updated feedforward compensation parameters or feedback control gain to generate more precise control commands to the tensioning actuator, thereby offsetting or mitigating tension transmission distortion caused by time-varying friction coefficients in real time, ensuring that the prestress of the steel wire wound onto the pole frame meets design requirements.

[0084] Example 2: Figure 2 A schematic diagram of the pole skeleton winding control system based on real-time tension feedback of the present invention is given. The pole skeleton winding control system based on real-time tension feedback includes the following modules:

[0085] The data acquisition module is used to acquire the wire tension feedback value monitored by the tension sensor in real time during the wire winding process, and simultaneously acquire the axial position information of the wire winding head on the pole frame.

[0086] The anomaly detection module is used to determine whether an anomaly in tension control has occurred based on the deviation between the wire tension feedback value and the preset tension setting value.

[0087] The state recognition module is used to continuously calculate the higher-order statistics of the wire tension feedback value within the sliding time window when an abnormality in tension control is detected, and to identify whether the friction state between the wire and the pole frame has changed direction based on the evolution trajectory of the higher-order statistics.

[0088] The feature association module is used to associate the evolution characteristics of higher-order statistics with the axial position information of the winding head when a change in direction occurs, so as to determine the spatial propagation characteristics of the change in friction state.

[0089] The distortion assessment module is used to assess the degree of tension transmission distortion caused by the time-varying friction coefficient, based on the spatial propagation characteristics of changes in friction state.

[0090] The compensation and adjustment module is used to dynamically compensate and adjust the parameters of the tension control system according to the degree of tension transmission distortion.

[0091] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0092] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0093] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and inventive constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0094] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0095] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0096] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0097] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for controlling the wire winding of a pole frame based on real-time tension feedback, characterized in that, Includes the following steps: S1. During the wire winding process, the wire tension feedback value monitored by the tension sensor is obtained in real time, and the axial position information of the wire winding head on the pole frame is obtained simultaneously. S2. Based on the deviation between the wire tension feedback value and the preset tension setting value, determine whether there is an abnormality in tension control; S3. When an abnormality in tension control is detected, the higher-order statistics of the wire tension feedback value are continuously calculated within the sliding time window, and the direction of the friction state between the wire and the pole frame is identified based on the evolution trajectory of the higher-order statistics. S4. When a change in direction occurs, the evolution characteristics of higher-order statistics are correlated with the axial position information of the winding head to determine the spatial propagation characteristics of the change in friction state. S5. Based on the spatial propagation characteristics of changes in friction state, assess the degree of tension transmission distortion caused by time-varying friction coefficient; S6. Based on the degree of tension transmission distortion, dynamically compensate and adjust the parameters of the tension control system. S1 includes: The wire tension feedback value monitored by the tension sensor is filtered in real time to obtain the wire tension feedback value used for control. The axial position information of the winding head is calibrated in real time to obtain the axial position information used for control. S2 includes: Calculate the real-time deviation between the wire tension feedback value used for control and the preset tension setting value; Calculate the moving range of this real-time deviation over a continuous time interval; The moving range is compared with the dynamic control limits established based on historical stable winding process data; When the range of movement continuously exceeds the dynamic control limit, it is judged that an abnormality in tension control has occurred. The dynamic control limit is dynamically generated based on the moving range statistical characteristics of the difference between the wire tension feedback value used for control and its corresponding preset tension setting value in historical stable winding process data. S3 includes: After determining that an abnormality in tension control has occurred, the skewness and kurtosis of the wire tension feedback value used for control are continuously calculated within the sliding time window. Using skewness and kurtosis as coordinates, an evolution trajectory within a sliding time window is constructed on a two-dimensional plane formed by the skewness axis and the kurtosis axis. Trajectory clustering analysis is performed on the points on the two-dimensional plane of the evolution trajectory to extract the movement direction and clustering pattern of the evolution trajectory; When the direction of movement of the evolution trajectory shows a trend of continuously moving away from the pre-defined stable friction state cluster center region, it is identified as a directional change in the friction state.

2. The method for controlling the wire winding of a pole frame based on real-time tension feedback according to claim 1, characterized in that, S4 includes: During the period when the friction state is identified as a directional change, the movement direction of the evolution trajectory and the time point corresponding to the aggregation pattern are recorded simultaneously, and the axial position sequence formed during this period is extracted for control purposes. Analyze the correlation between the rate of change of the direction of movement of the evolution trajectory and the changes in the axial position sequence; Based on the correlation, it is determined whether the change in friction state is a gradual propagation mode along the pole frame axis or a local concentrated propagation mode, which serves as a spatial propagation characteristic.

3. The method for controlling the wire winding of a pole frame based on real-time tension feedback according to claim 2, characterized in that, The correlation between the rate of change of the direction of movement of the evolution trajectory and the change of the axial position sequence is analyzed by calculating the Pearson correlation coefficient between the sequence of changes in the movement angle of the evolution trajectory on a two-dimensional plane composed of the skewness axis and the kurtosis axis and the amount of change in the axial position sequence.

4. The pole frame winding control method based on real-time tension feedback according to claim 3, characterized in that, S5 include: Based on the determined gradual propagation mode or localized centralized propagation mode, and combined with the movement direction and aggregation mode of the evolution trajectory, a model for assessing the degree of tension transmission distortion is established. Based on the axial position range corresponding to the spatial propagation characteristics and the degree of deviation of the evolution trajectory from the stable state, a comprehensive quantitative value of the degree of tension transmission distortion is calculated through the tension transmission distortion degree evaluation model.

5. The pole frame winding control method based on real-time tension feedback according to claim 4, characterized in that, S6 include: Based on the magnitude of the quantified value of the tension transmission distortion and the spatial propagation characteristics, a corresponding dynamic compensation adjustment amount is generated; When the spatial propagation characteristic is a progressive propagation mode, the feedforward compensation parameters of the tension control system are corrected by dynamic compensation adjustment. When the spatial propagation characteristic is a local centralized propagation mode, the feedback control gain of the tension control system is adjusted by the dynamic compensation adjustment amount.

6. A pole frame winding control system based on real-time tension feedback, used to implement the pole frame winding control method based on real-time tension feedback as described in any one of claims 1-5, characterized in that, Includes the following modules: The data acquisition module is used to acquire the wire tension feedback value monitored by the tension sensor in real time during the wire winding process, and simultaneously acquire the axial position information of the wire winding head on the pole frame. The anomaly detection module is used to determine whether an anomaly in tension control has occurred based on the deviation between the wire tension feedback value and the preset tension setting value. The state recognition module is used to continuously calculate the higher-order statistics of the wire tension feedback value within the sliding time window when an abnormality in tension control is detected, and to identify whether the friction state between the wire and the pole frame has changed direction based on the evolution trajectory of the higher-order statistics. The feature association module is used to associate the evolution characteristics of higher-order statistics with the axial position information of the winding head when a change in direction occurs, so as to determine the spatial propagation characteristics of the change in friction state. The distortion assessment module is used to assess the degree of tension transmission distortion caused by the time-varying friction coefficient, based on the spatial propagation characteristics of changes in friction state. The compensation and adjustment module is used to dynamically compensate and adjust the parameters of the tension control system according to the degree of tension transmission distortion.

Images