Laser broken line rapid identification method for spinning process

Abstract

The invention discloses a laser broken line rapid identification method for a spinning process, relates to the field of spinning detection, and is used for solving the problem of high identification error rate in a multi-material blending and frequent line changing environment. A multi-channel optical behavior characteristic matrix is formed by constructing a broken line recognition mechanism based on material perception and light intensity behavior characteristic collaborative modeling, collecting light intensity response data of yarn channels, constructing a dynamic detection sequence and extracting characteristics such as shielding amplitude gradient and recovery time delay. A yarn material identifier is introduced, a spectral response template is matched, a broken yarn judgment threshold group is generated, whether shielding sudden change and delay recovery characteristics are met or not is judged in combination with a detection sequence, a broken yarn event is accurately recognized, and recognition parameters are corrected in real time according to the shielding edge change rate and fluctuation stability in light intensity changes before and after yarn breakage. The dynamic optimization of the broken line judgment condition is realized, and the recognition precision and the response speed are improved.

Description

Technical Field

[0001] This invention relates to the field of textile inspection, and more specifically, to a method for rapid laser-based identification of broken threads in the textile process. Background Technology

[0002] Currently, the textile manufacturing industry is rapidly evolving towards automation, high speed, and flexible production. In particular, in typical processes such as knitting, warp knitting, and high-speed winding, hundreds or thousands of yarns need to be supplied continuously and stably. Once a yarn breaks, it can easily cause fabric defects or equipment downtime. To improve the real-time sensing capability of the production process, laser yarn breakage detection technology is gradually replacing traditional mechanical or photoelectric interruption detection methods. Existing laser detection systems typically use line laser irradiation or point laser beam passing methods. By sensing changes in light intensity at the receiving end, they can determine whether there is yarn obstruction, thereby achieving non-contact yarn breakage identification.

[0003] The shortcomings of existing technologies: In flexible weaving scenarios involving multi-variety blends and high-frequency yarn changes, the yarn materials, colors, and optical reaction characteristics are highly diverse. Existing laser yarn breakage recognition systems generally use a uniform recognition threshold or a fixed judgment model, failing to dynamically adapt to the optical characteristics of different yarns. This leads to a significant decrease in the recognition accuracy of the system in actual operation. Specifically, different yarn materials (such as nylon, cotton, and polyester) exhibit significant differences in reflectivity, absorptivity, and scattering behavior under laser irradiation. In particular, black or matte yarns, due to their strong absorption, are often misjudged as "broken" by the system, while highly reflective glossy yarns may still produce residual reflection signals after breaking and be misjudged as "unbroken." The system cannot dynamically adjust the recognition parameters according to material changes. Under complex operating conditions, the recognition system is forced to maintain a trade-off between sensitivity and false alarm rate, failing to guarantee timely response to yarn breakage and struggling to avoid false shutdowns caused by frequent false alarms. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the following solution is proposed to solve the problem of high identification error rate in the above-mentioned background technology under the conditions of multi-material blending and frequent line changing.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for rapid laser-based identification of broken threads in a textile process includes the following steps: The optical response signal of the yarn to the laser beam in operation is collected, and the light intensity response sequence is time-aligned to construct a dynamic detection sequence of light intensity changes. The light intensity occlusion features and transmission abrupt change features of each channel are extracted from the dynamic detection sequence. The occlusion amplitude gradient and occlusion recovery time delay parameters are calculated to generate a multi-channel optical behavior feature matrix. Collect the current yarn material identification information and select the corresponding type of material response parameter group from the preset spectral response template to establish the correspondence between yarn material and light intensity behavior characteristics; Based on the correspondence between yarn material and light intensity characteristics, determine the breakage judgment threshold group for each detection channel; The current detection sequence is compared with the disconnection judgment threshold group to determine whether the light intensity meets the combination characteristics of continuous occlusion and abrupt change mode. If it does, the corresponding detection channel is judged to have a disconnection event. The light intensity change segment within a fixed time period before and after the detection channel where the line breakage occurs is collected. Based on the change rate of the occlusion edge and the fluctuation stability of the change segment, the occlusion abruptness threshold and jump tolerance parameters in the material response parameter group are updated.

[0006] In a preferred embodiment, the optical response signal of the yarn to the laser beam during operation is acquired, and the light intensity response sequence is time-aligned to construct a dynamic detection sequence of multi-channel light intensity changes. Specific steps include: The yarn channel is irradiated at a fixed frequency by a laser emitting device; The light intensity change signal per unit time in each yarn channel is acquired by a photoelectric receiving device. The light intensity data is timestamped according to the time signal provided by the main controller. Continuous light intensity data is divided into fixed-length data segments and sorted to form a detection sequence.

[0007] In a preferred embodiment, the light intensity occlusion features and transmission abrupt change features of each channel are extracted from the dynamic detection sequence, the occlusion amplitude gradient and occlusion recovery delay parameters are calculated, and a dynamic detection sequence of light intensity changes is constructed, including the following steps: Traverse the dynamic detection sequence corresponding to each channel, identify the segment in the time-continuous data segment where the light intensity value continuously decreases from high to low and continues to exceed the preset occlusion amplitude threshold, and mark it as the start time of the occlusion event. Determine whether the light intensity value has an upward trend after it drops to a minimum value, and mark the time when the value recovers to the initial stable level range as the end time of the occlusion event. Calculate the time interval between the start and end times as the occlusion recovery delay parameter. Between the start of the occlusion event and the minimum value, the pair of points with the fastest decrease in light intensity is identified. The occlusion rate is calculated by the ratio of the difference in light intensity decrease between the pair of points to the sampling time interval. This calculation result is used as the occlusion amplitude gradient. Record the occlusion recovery delay parameter and occlusion amplitude gradient value extracted from each data segment to form a set of light intensity occlusion feature parameters for the data channel. Arrange the light intensity occlusion feature parameter sets of all channels in chronological order to construct a dynamic detection sequence of light intensity changes.

[0008] In a preferred embodiment, the occlusion recovery delay parameter and occlusion amplitude gradient value extracted from each data segment are recorded to form a set of light intensity occlusion feature parameters for the data channel, including the following steps: In the dynamic detection sequence corresponding to each detection channel, each divided data segment is traversed sequentially, and the occlusion recovery delay parameter and occlusion amplitude gradient value are extracted in each data segment. For each data segment, when a complete occlusion event is successfully identified, the occlusion recovery delay parameter corresponding to the occlusion event is stored as a time length value and bound to the data segment start timestamp as the first field of a record unit. Record the occlusion magnitude gradient value as the second field of the recording unit; The first field and the second field are combined to form a binary feature record unit, which is then associated with the data segment number and the detection channel identification information to form a set of light intensity occlusion feature parameters for the detection channel.

[0009] In a preferred embodiment, current yarn material identification information is collected, and a corresponding type of material response parameter group is selected from a preset spectral response template to establish a correspondence between yarn material and light intensity behavior characteristics. Specific steps include: The yarn material identifier can be entered from the user interface or automatically read from the loom control system. Search the preset template library for the laser reflection parameters, light absorption parameters, and historical occlusion response curve parameters corresponding to the material; The found material parameters are calibrated with the current channel optical behavior feature matrix to generate channel material adaptation relationships.

[0010] In a preferred embodiment, based on the correspondence between yarn material and light intensity characteristics, a set of breakage judgment thresholds for each detection channel is determined, including the following steps: The optical absorption intensity of the material is used as the adjustment benchmark for the threshold of shading abrupt change intensity; Set the upper limit of the distribution of occlusion recovery time under the same material in history as the critical time for light intensity recovery; The maximum stable amplitude of the occlusion event is used as the jump tolerance parameter. The occlusion sudden change intensity threshold, the light intensity recovery critical time, and the jump tolerance parameter together constitute the disconnection judgment threshold group of the detection channel.

[0011] In a preferred embodiment, the current detection sequence is compared with a set of disconnection judgment thresholds to determine whether the light intensity meets the characteristics of a combination of continuous interruption and abrupt change modes. If it does, a disconnection event is determined to have occurred in the corresponding detection channel, including the following steps: The magnitude of occlusion changes in the detected sequence is compared with the mutation intensity threshold to determine whether the mutation intensity threshold is exceeded. Determine if the duration of occlusion exceeds the critical recovery delay time; If the occlusion range exceeds the occlusion range threshold and the recovery time is greater than the recovery time threshold, then the disconnection mode characteristics are met, and a disconnection event marker is generated.

[0012] In a preferred embodiment, the disconnection event marker includes the disconnection channel number, the timestamp of the disconnection, and the start and end index values ​​of the light intensity change segment before and after the disconnection in the detection sequence.

[0013] In a preferred embodiment, the light intensity change segment of the detection channel where the line breakage occurs within a fixed time period before and after the determination is collected. Based on the occlusion edge change rate and fluctuation stability of the change segment, the occlusion abruptness threshold and jump tolerance parameters in the material response parameter group are updated, including the following steps: Extract light intensity variation data of the broken channel within a symmetrical time interval before and after the breakage; The slope of the occlusion edge in the data segment is statistically summarized, and its maximum fluctuation on the time axis is measured. The light intensity variation data and the maximum fluctuation degree are used to adjust the original abrupt change intensity threshold and jump tolerance parameters, and then updated with the original parameters using a weighted method.

[0014] In a preferred embodiment, the weighting method used in the update process is such that the historical disconnection frequency of the channel is positively correlated with the parameter update weight.

[0015] The technical effects and advantages of the laser-based rapid thread breakage identification method for textile processes according to the present invention are as follows: This invention achieves rapid identification and dynamic adaptation of multi-channel laser-based yarn breakage during textile manufacturing by constructing a yarn breakage identification mechanism based on collaborative modeling of material perception and light intensity behavior characteristics. By collecting light intensity response data of the yarn channels during operation, a dynamic detection sequence of light intensity changes is constructed. Key features such as occlusion amplitude gradient and recovery delay are extracted to form a multi-channel optical behavior feature matrix. Yarn material identification parameters are introduced, and combined with a preset spectral response template, a mapping relationship between material and behavior features is established, generating a set of breakage judgment thresholds that match the current operating state. Furthermore, by combining and matching the light intensity detection sequence with the breakage judgment threshold group, it can identify whether the conditions of occlusion mutation and delayed recovery are met simultaneously, accurately locate the breakage event, and dynamically correct the mutation intensity threshold and jump tolerance parameters by measuring the change rate and fluctuation stability of the occlusion edge in the light intensity change segment before and after the breakage, thereby continuously optimizing the breakage judgment conditions. This significantly improves the breakage recognition accuracy and response speed under conditions of yarn material changes and frequent operational disturbances, effectively avoiding misjudgment and missed judgment caused by fixed parameter settings, and providing a stable and low-latency breakage detection method for textile production. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a laser-based rapid thread breakage identification method for textile processes according to the present invention. Detailed Implementation

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] In order to achieve the above objectives, Figure 1 A schematic diagram of the structure of a laser-based rapid thread breakage identification method for textile processes according to the present invention is provided, which specifically includes the following steps; The optical response signal of the yarn to the laser beam in operation is collected, and the light intensity response sequence is time-aligned to construct a dynamic detection sequence of light intensity changes. The light intensity occlusion features and transmission abrupt change features of each channel are extracted from the dynamic detection sequence. The occlusion amplitude gradient and occlusion recovery time delay parameters are calculated to generate a multi-channel optical behavior feature matrix. Collect the current yarn material identification information and select the corresponding type of material response parameter group from the preset spectral response template to establish the correspondence between yarn material and light intensity behavior characteristics; Based on the correspondence between yarn material and light intensity characteristics, determine the breakage judgment threshold group for each detection channel; The current detection sequence is compared with the disconnection judgment threshold group to determine whether the light intensity meets the combination characteristics of continuous occlusion and abrupt change mode. If it does, the corresponding detection channel is judged to have a disconnection event. The light intensity change segment within a fixed time period before and after the detection channel where the line breakage occurs is collected. Based on the change rate of the occlusion edge and the fluctuation stability of the change segment, the occlusion abruptness threshold and jump tolerance parameters in the material response parameter group are updated.

[0019] Step 1: Acquire the optical response signal of the yarn to the laser beam while it is in operation, and time-align the light intensity response sequence to construct a dynamic detection sequence for light intensity changes. The specific implementation is as follows: During the yarn operation, the target yarn area is stably illuminated by a laser emitting device. Each laser beam illuminates the corresponding yarn channel at a fixed emission frequency, so that the beam forms a continuous illumination state as it passes through the yarn path. Each yarn channel is equipped with a corresponding light intensity acquisition path to obtain the light intensity response signal when the light passes through or is blocked.

[0020] As the yarn moves, it may obstruct or penetrate the laser beam. A photodetector senses the remaining intensity of the laser after it passes through the yarn, recording the light intensity value at each time point. The light intensity sampling frequency is set according to the laser emission frequency and the controller's data processing capability, typically a fixed value to ensure timing consistency. To ensure data synchronization across different channels under the same time reference, each set of light intensity values ​​collected by the photodetector is aligned with a unified clock signal provided by the main controller. Specifically, for each light intensity sample value, its sampling time is compared with the main controller clock, and a corresponding timestamp is added. This timestamp characterizes the position of the data point in the global time series, ensuring comparability and synchronization of subsequent multi-channel data during processing.

[0021] After the timestamp calibration is completed, the light intensity data of each channel are archived in chronological order and divided into multiple data segments at fixed time intervals. Each data segment contains several continuous and equally spaced sampling points, representing the response behavior of the yarn to the laser within a continuous time period. The time length is set by the system and can be adjusted according to the actual working conditions, such as 100 milliseconds, 200 milliseconds, etc.

[0022] After the data segments are divided, they are organized in chronological order to form the dynamic detection sequence of the detection channel. Each yarn channel corresponds to a dynamic detection sequence. The data set composed of multiple detection channels is the dynamic detection sequence of multi-channel light intensity change. This dynamic detection sequence will serve as the basic input data for subsequent occlusion feature extraction and thread breakage event judgment.

[0023] Step 2: Extract the light intensity occlusion features and transmission abrupt change features of each channel from the dynamic detection sequence, calculate the occlusion amplitude gradient and occlusion recovery delay parameters, and generate a multi-channel optical behavior feature matrix. The specific implementation is as follows: The dynamic detection sequence corresponding to each channel is traversed. Each dynamic detection sequence consists of a set of light intensity data segments sampled according to a unified time base. Each data segment represents a sampling window of fixed time length, recording the continuous state of light intensity value changes over time. The data segments of each channel are traversed in turn, and the trend of light intensity change is analyzed segment by segment. Based on the changes in light intensity values ​​in each data segment, it is identified whether the light intensity value continuously decreases from a high level. If the light intensity decrease continues for multiple sampling points and the cumulative decrease exceeds the preset occlusion amplitude threshold, then the segment is considered the beginning of an occlusion event. The starting time of the occlusion event is the timestamp corresponding to the sampling point where the light intensity first drops significantly and exceeds the occlusion amplitude threshold. The occlusion amplitude threshold is defined as: the average value of the stable light intensity range of each channel under normal operating conditions minus a preset deviation value. The deviation value is obtained based on the statistics of historical operating samples and is used to eliminate non-occlusion fluctuations caused by environmental disturbances.

[0024] The occlusion event termination time and occlusion recovery delay parameter are determined by continuously monitoring the change in light intensity value. After it drops to a minimum value, if a continuous upward trend appears and it recovers to the fluctuation range of the average light intensity before the start in multiple sampling points, the occlusion event is considered to have ended. The occlusion termination time is the sampling time point when the light intensity first recovers to the stable range. The occlusion recovery delay parameter is obtained by subtracting the termination time timestamp from the start time timestamp of the occlusion event to obtain the duration of the occlusion event, which is used to quantify the duration of the occlusion phenomenon.

[0025] The occlusion magnitude gradient is calculated by comparing the light intensity differences between all adjacent sampling points in the light intensity data from the start time to the minimum value in the same occlusion event, identifying the pair of points with the fastest light intensity decrease, and then comparing the light intensity decrease between the two sampling points with their time interval. This yields the maximum light intensity decrease rate per unit time in the occlusion event, which is the occlusion magnitude gradient. The occlusion magnitude gradient reflects the degree of abrupt change in occlusion behavior. The larger the value, the more rapid the change in yarn response to laser occlusion, and it is often strongly correlated with the early characteristics of yarn breakage.

[0026] The process of constructing the light intensity occlusion feature parameter set involves traversing the dynamic detection sequence of each detection channel. Each channel corresponds to a dynamic detection sequence, which consists of a set of data segments arranged in chronological order. Each data segment contains light intensity values ​​sampled continuously over a period of time, with a fixed sampling start timestamp and length. During the traversal, these data segments are read one by one to facilitate subsequent identification of occlusion behavior and parameter extraction. In each data segment, identify whether there is a complete occlusion event. An occlusion event must meet the following conditions: the light intensity value shows a significant downward trend, the decrease exceeds the preset occlusion amplitude threshold, and there is a significant rebound after the decrease, eventually returning to the range near the initial stable value. If the above conditions are met, the data segment is considered to contain a complete occlusion event. After identifying the complete occlusion event, feature parameters are extracted as follows: For data segments from which the occlusion recovery delay parameter and occlusion amplitude gradient value are successfully extracted, the system records them as a standardized binary feature recording unit. The first field is the occlusion recovery delay parameter, which represents the duration of the occlusion event; the second field is the occlusion amplitude gradient value, which represents the abrupt rate of light intensity decrease during the occlusion process. This recording unit is bound to the sampling start timestamp of the current data segment and the position number of the data segment in the current channel detection sequence (i.e., the data segment number), and is appended with the unique identifier of the detection channel.

[0027] The binary feature record units extracted from all data segments within each channel are arranged in chronological order to form a set of light intensity occlusion feature parameters for that channel. This set is stored in a time-series structure to ensure that each feature record can be retrieved and analyzed according to the actual sampling order. Each record can be traced back to its corresponding channel number and data segment number, ensuring traceability and structural consistency.

[0028] The light intensity occlusion feature parameters of all channels are collected and merged in chronological order. Within the same time slice, the records of each channel are aligned and arranged to form a matrix structure. Each row of this matrix corresponds to a time window, and the columns correspond to the occlusion recovery delay parameters and occlusion amplitude gradient values ​​of each channel within that window, forming a structured optical behavior feature matrix for subsequent line break identification and pattern extraction algorithms.

[0029] Collect the current yarn material identification information, select the corresponding type of material response parameter group from the preset spectral response template, and establish the correspondence between yarn material and light intensity behavior characteristics, including the following steps: The current yarn material identification information is collected in two ways: one is for the operator to manually input the material type identification of the currently loaded yarn through the interactive interface, such as "polyester", "cotton", "nylon", etc.; the other is to automatically collect the yarn material identification field through the status reading interface of the loom control system. The system extracts the yarn material identification information corresponding to the current channel from the control command or formula parameters and automatically completes the identification input. The above yarn material identification information adopts a standard format definition, such as using strings or coded numbers to ensure consistency with the index structure in the template library. The material response parameter set is searched in the preset template library. This library contains a preset yarn optical response template, which is stored according to yarn material type. Each material type is associated with a set of response parameters formed by historical experimental data. These parameters mainly include: laser reflection parameters, which represent the reflectivity level of the material to a specific wavelength of laser light and are defined as the ratio of incident light power to reflected light power; light absorption parameters, which represent the material's ability to absorb laser energy and affect the dynamic changes in light intensity after transmission; and occlusion response curve parameters, which indicate the typical light intensity decrease trend curve caused by the material under standard occlusion conditions. Based on the material identifier obtained in the preceding steps, the matching mechanism is used to search for the corresponding material response parameter group in the template library. If the search is successful, the set of parameters is extracted and prepared for calibration with the current detection data. If no complete match is found, the default fallback mechanism can be triggered, and the general material template group is used as a temporary adaptation parameter, but a prompt is given on the interface.

[0030] After completing the collection of yarn material identification information and extraction of material response parameter sets, it is necessary to perform parameter mapping and binding of the optical response characteristics of the material with the optical behavior feature matrix generated in each detection channel. The core objective of this process is to establish a one-to-one correspondence between the actual observed behavior of the detection channel and the theoretical material characteristics, so as to achieve personalized adjustment of the discrimination benchmark. First, according to the channel identifier, the corresponding light intensity behavior feature matrix is ​​located. This matrix consists of light intensity occlusion recovery delay parameters and occlusion amplitude gradient values ​​arranged in chronological order, which are used to reflect the optical dynamic response of the yarn during operation. The yarn material identifier bound to the current channel is called, and the previously extracted laser reflection parameters, light absorption parameters and occlusion response curves are retrieved. A set of mapping indexes is established to form the binding fields between the channel number, material number and material response parameters. After establishing the mapping relationship, material response parameters are injected as correction factors into the preprocessing module of the feature matrix, enabling subsequent analysis and identification of light intensity behavior characteristics to be material-adaptive. For example, if the light absorption parameter of the current material is detected to be high, the occlusion amplitude gradient value in the feature matrix will be normalized according to the preset absorption and transmission correction rules. Furthermore, if the historical occlusion response curve of the current material exhibits a gradual downward slope, the identification threshold for occlusion recovery delay in the judgment logic will be automatically adjusted to avoid misjudgments caused by differences in material light absorption capabilities. The entire mapping and binding process is completed by the system during the initialization phase and remains continuously effective throughout the entire line break identification process, ensuring that the judgment logic has material perception capabilities, thereby achieving personalized, accurate, and stable identification strategies. Through this mapping calibration mechanism, even under complex working conditions such as multi-material blending and frequent yarn changes, high judgment sensitivity and false alarm suppression capabilities can still be maintained.

[0031] Suppose a loom is currently processing white nylon yarn. The operator selects "Material Type: Nylon" on the loom control interface and immediately retrieves three parameters corresponding to the nylon yarn from a preset spectral response template library: medium to high laser reflectivity, low light absorption, and a historical occlusion response curve showing small abrupt changes in occlusion but slow recovery. Simultaneously, light intensity variation data for each yarn channel has been acquired from the photoelectric receiving device, and an optical behavior feature matrix for that channel has been generated. For example, in a certain channel, an event record shows a long occlusion recovery time but a small decrease in light intensity.

[0032] At this point, the optical behavior matrix of the channel will be compared one-to-one with the typical response characteristics of nylon yarn. For example, based on the combination trend of "small occlusion amplitude and large recovery delay", the system confirms that this is highly matched with the material response curve of "strong reflection, weak absorption and slow light spot recovery of nylon yarn". Therefore, the current behavior is marked as "normal material occlusion" and no disconnection alarm is triggered. However, if this type of feature matrix appears on cotton yarn or wool material with "high light absorption and fast occlusion recovery", it will be judged as an abnormal occlusion event and a warning or disconnection prompt will be issued.

[0033] This example illustrates that for the same light intensity change behavior, different materials have different interpretation thresholds. By pre-binding material characteristic parameters, false alarms and false negatives can be minimized.

[0034] Based on the correspondence between yarn material and light intensity characteristics, the threshold group for determining yarn breakage in each detection channel is determined. The specific steps include: After identifying the yarn material used in the current weaving process, for example, determining that polyester yarn is being used, the optical absorption intensity parameter corresponding to the material is first extracted from the optical response template library. The optical absorption intensity parameter refers to the strength of the yarn material's ability to absorb the energy of a laser beam, usually derived from reflectivity and transmittance. It is used to characterize the magnitude of light intensity change when the laser beam passes through the yarn. This optical absorption intensity parameter is used as the adjustment benchmark for the occlusion abrupt change intensity threshold. Based on the standard of the basic abrupt change threshold, it is fine-tuned according to the level of absorption intensity to adapt to the occlusion response characteristics of different materials. The higher the optical absorption intensity, the greater the decrease in light intensity caused by the occlusion abrupt change, and the more lenient the corresponding abrupt change intensity threshold is set; conversely, for materials with lower absorption intensity, a more stringent abrupt change detection threshold needs to be set to ensure recognition sensitivity.

[0035] Subsequently, the distribution of shading recovery time recorded in the historical normal operation data of the material is called. That is, the time range required for the light intensity to recover after the yarn of the material was blocked in a non-broken state in history. By statistically analyzing the shading recovery delay parameters in these historical records, the upper limit of its statistical distribution is determined, and this upper limit is set as the light intensity recovery critical time of the current material. The light intensity recovery critical time is the time limit used to judge whether the light intensity has recovered to the normal state. If it has not recovered after exceeding the time limit, it is initially judged that there is a risk of yarn breakage. Next, the maximum stable occlusion amplitude of typical occlusion events under this material in historical data is calculated, that is, the most severe light intensity occlusion degree in the non-disconnection state. This is used to define the jump tolerance parameter of the current material. The jump tolerance parameter is used to distinguish between critical occlusion events and abnormal disconnection events. When the light intensity drop of a certain occlusion event exceeds this jump tolerance value, a further disconnection judgment process will be triggered.

[0036] Finally, the above three parameters—namely, the occlusion mutation intensity threshold (adjusted according to absorption intensity), the light intensity recovery critical time (derived from the upper limit of historical occlusion recovery distribution), and the jump tolerance parameter (taken from the material's historical maximum occlusion amplitude)—are combined to form the current channel's disconnection judgment threshold group under the material conditions. Each detection channel is indexed and bound with its own channel number and material identifier to ensure that the disconnection judgment logic has material dependence and channel-specific adjustment capabilities.

[0037] The breakage judgment threshold group is written into the detection parameter control list for subsequent breakage recognition logic based on light intensity change trend, so as to achieve high-precision breakage judgment with material adaptation. By constructing such a multi-parameter threshold group based on the optical properties of the material, the problem of misjudgment and missed judgment caused by different yarn materials can be effectively avoided, ensuring that the breakage behavior can still be stably and accurately identified in complex weaving environment.

[0038] The current detection sequence is compared with the disconnection judgment threshold group to determine whether the light intensity meets the combined characteristics of continuous occlusion and abrupt change mode. If it does, a disconnection event is determined to have occurred in the corresponding detection channel, including the following steps: After constructing the dynamic detection sequence for light intensity changes, a breakage identification process is executed on the data sequence in each detection channel. This process compares the occlusion events recorded in the current channel one by one based on the breakage judgment threshold group previously established according to the yarn material. The specific process is as follows: From the dynamic detection sequence corresponding to the detection channel, data segments marked as complete occlusion events are read one by one. Each data segment contains two key feature parameters: occlusion recovery delay parameter and occlusion amplitude gradient value. The occlusion recovery delay parameter refers to the time required from the start of the occlusion event to the light intensity recovering to the initial stable range; the occlusion amplitude gradient value is a characterization of the rate of change during the light intensity decrease, reflecting the degree of abrupt change when occlusion occurs. The occlusion amplitude gradient value of the data segment is compared with the occlusion abrupt change intensity threshold configured in its detection channel to determine whether the light intensity change has sufficient abrupt change characteristics. If the gradient value is equal to or exceeds the occlusion abrupt change intensity threshold, it is considered to meet the requirements of occlusion abrupt change intensity. It is then determined whether the occlusion recovery delay parameter recorded in the data segment exceeds the light intensity recovery critical time corresponding to the current yarn material. This critical time is the maximum allowable delay for the yarn to complete light intensity recovery under non-broken conditions, set based on historical data, and is used to identify recovery failures caused by potential yarn breakage. Only when both of the above conditions are met simultaneously, namely, the intensity of the occlusion change exceeds the threshold and the light intensity recovery delay exceeds the critical time, is the occlusion event considered to be a broken yarn pattern feature. It is then assumed that the yarn corresponding to the channel has broken within the detection time period. At this time, a broken yarn event marker is generated under the channel identification information, and the occurrence time of the event, channel number, broken yarn trigger feature value, etc. are packaged and written into the broken yarn event record list for subsequent alarm response, loom shutdown processing, or yarn replacement operation.

[0039] After the system detects an occlusion event that meets the characteristics of a disconnection pattern, it generates a disconnection event marker. This marker is used to accurately record and process the disconnection behavior. The disconnection event marker mainly consists of three core fields: the disconnection channel number, the timestamp of the disconnection, and the start and end index values ​​(the position of the light intensity change segment in the detection sequence). The broken yarn channel number is a unique identifier for the current detection channel within the system, used to identify the physical channel where the broken yarn event occurred. This number is usually encoded as an integer, such as "CH_03" representing detection channel number 3, facilitating a one-to-one correspondence with the laser receiving channel and yarn position. In subsequent alarm linkage or data analysis, this number can be used to directly trace the location of the problem. The timestamp of the thread breakage event records the exact moment the event occurred, specifically the absolute time stamp applied by the detection system the instant the thread breakage characteristic is identified. The timestamp is typically synchronized with the system's main controller clock signal and can be formatted using the standard "year-month-day hour:minute:second" structure. For example, "2025-07-09 11:32:17", to ensure accurate alignment with subsequent loom operation logs or alarm records. The start and end index values ​​refer to the index numbers of the starting and ending data segments of the light intensity change process involved in this disconnection event in the dynamic detection sequence of the current detection channel. The start index value is the position number of the data segment in the sequence when the occlusion event begins; the end index value is the data segment number corresponding to the end of the occlusion event when it is recovered or judged as disconnection. These two index values ​​together define the coverage range of the disconnection behavior in the complete detection sequence, which facilitates the subsequent review or pattern analysis of the light intensity evolution process before and after the disconnection. The index number is an auto-incrementing sequence maintained internally by the system, generally starting from 0, and automatically increasing with each new data segment.

[0040] The detection channel that detects a broken line collects light intensity changes within a fixed time period before and after the detection. Based on the rate of change of the occlusion edge and the stability of the fluctuation during the change segment, the occlusion abruptness threshold and jump tolerance parameters in the material response parameter group are updated, including the following steps: After detecting a disconnection event and generating a disconnection event marker, the system automatically performs post-disconnection analysis and parameter correction operations. It collects light intensity change data of the detection channel where the disconnection occurred within a fixed time interval before and after the event, and corrects the occlusion abruptness threshold and jump tolerance parameters in the stored material response parameter group based on the occlusion edge change rate and fluctuation stability. The first step is to extract light intensity change data within a symmetrical time period before and after the line break. Based on the time stamp and channel number recorded in the line break event marker, the light intensity detection sequence of the corresponding channel is located. Within this sequence, with the line break time point as the center, equal-length time windows (e.g., 200 milliseconds before and after) are taken, and all light intensity sampling data within this time period are extracted. This data segment covers the entire process of light intensity evolution before and after the line break, including both the occlusion formation process before the line break and the possible reflective interference or material fluctuation characteristics after the line break. The second step involves statistically analyzing the rate of change and fluctuation stability of the occlusion edge. Within the extracted data segment, occlusion edge analysis is performed. First, the light intensity decrease edge segment is identified, and its descent slope on the time axis is calculated, representing the magnitude of the light intensity change per unit time at the steepest point. Then, the range of light intensity changes across the entire data segment is analyzed, calculating the maximum fluctuation amplitude, which is the difference between the local maximum and minimum values ​​of the light intensity signal. This reflects the severity of occlusion and the level of signal stability under actual working conditions. The descent slope reflects abrupt changes, while the maximum fluctuation amplitude reflects tolerance requirements. The third step is to update the threshold parameters in the material response parameter group. The slope of the currently obtained occlusion edge is used as the reference value for the abrupt change intensity, and the maximum fluctuation amplitude is used as the reference value for the jump tolerance. These are then weighted and fused with the occlusion abrupt change threshold and jump tolerance parameters already set in the current material response parameter group. The weighted update method is to combine the original parameter value with the current statistical value according to a set ratio (e.g., 70% retaining historical experience values ​​and 30% introducing new data) to avoid interference from outliers while gradually adapting to actual working conditions. For example, if the original jump tolerance parameter for the current material is 20 units of light intensity, and the current fluctuation amplitude is 26 units of light intensity, then the updated tolerance parameter can be set as a weighted average of the original and new values ​​(e.g., 0.7×20+0.3×26=21.8), and stored back in the parameter group for the current yarn material. The weighting method is set as follows: the higher the historical disconnection frequency of the detection channel, the greater its parameter update weight and the higher its participation in the update; the lower the disconnection frequency, the more stable the parameters tend to remain. This is based on the following technical logic: If a channel frequently experiences disconnection events, the original parameter settings may be inaccurate and need to be adapted more quickly. If a channel is operating stably, frequent parameter updates may introduce unnecessary disturbances. Therefore, a dynamic weight update mechanism is used for differentiated learning.

[0041] In actual updates, the following weighted logic is used for the light intensity mutation value and recovery delay value generated by each disconnection event: First, extract the new occlusion mutation intensity and recovery delay value for the current disconnection event; count the frequency of disconnection events in the most recent N detections of this detection channel and set it as the historical disconnection frequency factor; based on the magnitude of this factor, the system looks up its corresponding update weight from the parameter library (e.g., by table lookup or segmented setting), and uses this weight to perform weighted fusion of the newly extracted parameter value and the current parameter value. If the frequency is high, the new value has a large weight and the update magnitude is large; if the frequency is low, the original value remains dominant and the update magnitude is small. For example, if the historical disconnection frequency of the current channel is 5%, the corresponding update weight obtained by looking up the table is 0.7; the original parameter value is 50, and the new observation value is 55, then the updated parameter is "a weighted value between 50 and 55 that is close to 55", with the value close to the new observation value.

[0042] This process does not change the core threshold rules of the wire breakage judgment logic. It only makes minor adjustments based on the current material performance after each wire breakage event. It has the ability to learn gradually, which helps to maintain the accuracy and stability of wire breakage detection when there are frequent fluctuations in multiple batches of materials or drastic changes in production process conditions. The updated parameter set will be called in subsequent wire breakage judgment processes to form a material adaptation closed-loop path.

[0043] The rate of change of occlusion edge is a parameter used to quantify the speed of change of the laser received light intensity signal during the occlusion process of the yarn. It describes the degree of abrupt change in the light intensity signal at the initial stage of occlusion and is obtained based on a continuous light intensity value sequence collected in the detection channel through the following process: To identify occlusion edge segments, in each data segment related to thread breakage determination, identify the continuous decreasing range of light intensity from a stable state until reaching a local minimum; these are called occlusion edge segments. Extract the most drastic point pairs: within the occlusion edge segments, find the two adjacent sampling points with the most drastic decrease in light intensity, i.e., the point with the largest light intensity difference between the two time points; these two points are called change point pairs. Divide the light intensity difference between the two sampling points by the time interval between them to obtain the light intensity decrease per unit time; this value is the occlusion edge change rate. The larger the occlusion edge change rate, the faster the yarn occlusion occurs, and the more significant the sudden drop in light intensity signal. This is usually used to determine whether the occlusion belongs to the category of sudden thread breakage behavior.

[0044] Fluctuation stability is a parameter used to reflect whether the intensity variation range of a laser signal remains relatively stable within a fixed time interval, or whether there are drastic fluctuations. It is used to measure the stability of the signal after obstruction. The method of obtaining it is to extract the light intensity data within a symmetrical time period before and after the wire breakage event based on the wire breakage event timestamp, identify all local maxima and minima within this time period, i.e., the peak and valley values ​​of the light intensity, and find the difference between the maximum maximum and minimum values ​​within this time interval, which is the maximum fluctuation amplitude of this segment. If the maximum fluctuation amplitude is less than the jump tolerance parameter set by the system for this yarn material, the light intensity fluctuation of this segment is considered stable; otherwise, the fluctuation is considered unstable, and there may be interference or abnormal reflection. Therefore, fluctuation stability is not a single numerical value, but a judgment result based on whether the fluctuation amplitude is within the tolerable range. In implementation, the fluctuation amplitude value itself can be used as a numerical representation of this parameter and used for threshold adjustment and model correction.

[0045] After the threshold group update is completed, the following recognition operation will be performed on each data segment in the current detection sequence: The occlusion amplitude and recovery delay extracted from the data segment are evaluated. The occlusion amplitude is compared with the mutation intensity threshold in the threshold group to determine whether it exceeds the threshold. The recovery delay is determined to be greater than the recovery time threshold. If both conditions are met, it is determined that the channel has experienced a stable and irreversible laser occlusion behavior during the period. The channel number, event occurrence time and corresponding data segment index value are immediately packaged into a broken thread event marker and a broken thread identification signal is sent to the upper-level control module or the loom control system. The rapid identification of laser-induced line breaks relies on the cooperation of the following three aspects: a sliding detection mechanism based on a fixed sampling window, where data enters the identification process in real time without batch processing; a real-time parameter update mechanism, which ensures that the judgment criteria always match the current operating status, avoiding secondary confirmation or multiple misjudgments; and a structured storage and indexing mechanism for feature parameters, which can directly locate the time and channel position of the line break event, thereby achieving millisecond-level response and accurate positioning.

[0046] It should be noted that the historical disconnection frequency refers to the ratio of the number of times a certain detection channel is identified as having a disconnection event within a set statistical period to the total number of samples taken within that period. For example, if a channel detects 10 disconnections in the past 1000 detections, its historical disconnection frequency is 1%. Parameter update weight refers to the contribution of the newly acquired parameter value to the weighted update process when updating key indicators in the material response parameter group, such as the occlusion mutation threshold and jump tolerance parameter. The higher the weight, the greater the impact of the new value on the final result.

[0047] It should be noted that the threshold information in this embodiment was set in advance by professionals and will not be explained in detail here. Some parameters in the embodiment may have the same English letters, but they are explained with different meanings when used, and will not be explained one by one here.

[0048] This invention achieves rapid identification and dynamic adaptation of multi-channel laser-based yarn breakage during textile manufacturing by constructing a yarn breakage identification mechanism based on collaborative modeling of material perception and light intensity behavior characteristics. By collecting light intensity response data of the yarn channels during operation, a dynamic detection sequence of light intensity changes is constructed. Key features such as occlusion amplitude gradient and recovery delay are extracted to form a multi-channel optical behavior feature matrix. Yarn material identification parameters are introduced, and combined with a preset spectral response template, a mapping relationship between material and behavior features is established, generating a set of breakage judgment thresholds that match the current operating state. Furthermore, by combining and matching the light intensity detection sequence with the breakage judgment threshold group, it can identify whether the conditions of occlusion mutation and delayed recovery are met simultaneously, accurately locate the breakage event, and dynamically correct the mutation intensity threshold and jump tolerance parameters by measuring the change rate and fluctuation stability of the occlusion edge in the light intensity change segment before and after the breakage, thereby continuously optimizing the breakage judgment conditions. This significantly improves the breakage recognition accuracy and response speed under conditions of yarn material changes and frequent operational disturbances, effectively avoiding misjudgment and missed judgment caused by fixed parameter settings, and providing a stable and low-latency breakage detection method for textile production.

[0049] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0050] 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.

[0051] 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 design 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.

[0052] 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.

[0053] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0054] 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 rapid laser thread breakage identification in textile processes, characterized in that: The steps include: The optical response signal of the yarn to the laser beam in operation is collected, and the light intensity response sequence is time-aligned to construct a dynamic detection sequence of light intensity changes. The light intensity occlusion features and transmission abrupt change features of each channel are extracted from the dynamic detection sequence. The occlusion amplitude gradient and occlusion recovery time delay parameters are calculated to generate a multi-channel optical behavior feature matrix. Collect the current yarn material identification information and select the corresponding type of material response parameter group from the preset spectral response template to establish the correspondence between yarn material and light intensity behavior characteristics; Based on the correspondence between yarn material and light intensity characteristics, determine the breakage judgment threshold group for each detection channel; The current detection sequence is compared with the disconnection judgment threshold group to determine whether the light intensity meets the combination characteristics of continuous occlusion and abrupt change mode. If it does, the corresponding detection channel is judged to have a disconnection event. The light intensity change segment within a fixed time period before and after the detection channel where the line breakage occurs is collected. Based on the change rate of the occlusion edge and the fluctuation stability of the change segment, the occlusion abruptness threshold and jump tolerance parameters in the material response parameter group are updated.

2. The laser-based rapid thread breakage identification method for textile processes according to claim 1, characterized in that: The optical response signal of the yarn to the laser beam during operation is acquired, and the light intensity response sequence is time-aligned to construct a dynamic detection sequence of multi-channel light intensity changes, including the following steps: The yarn channel is irradiated at a fixed frequency by a laser emitting device; The light intensity change signal per unit time in each yarn channel is acquired by a photoelectric receiving device. The light intensity data is timestamped according to the time signal provided by the main controller. Continuous light intensity data is divided into fixed-length data segments and sorted to form a dynamic detection sequence.

3. The method for rapid laser thread breakage identification in the textile process according to claim 2, characterized in that: Light intensity occlusion features and transmission abrupt change features of each channel were extracted from the dynamic detection sequence. Calculate the occlusion magnitude gradient and occlusion recovery delay parameters to generate a multi-channel optical behavior feature matrix, including the following steps: Traverse the dynamic detection sequence corresponding to each channel, identify the segment in the time-continuous data segment where the light intensity value continuously decreases from high to low and continues to exceed the preset occlusion amplitude threshold, and mark it as the start time of the occlusion event. Determine whether the light intensity value has an upward trend after it drops to a minimum value, and mark the time when the value recovers to the initial stable level range as the end time of the occlusion event. Calculate the time interval between the start and end times as the occlusion recovery delay parameter. Between the start of the occlusion event and the minimum value, the pair of points with the fastest decrease in light intensity is identified. The occlusion rate is calculated by the ratio of the difference in light intensity decrease between the pair of points to the sampling time interval. This calculation result is used as the occlusion amplitude gradient. Record the occlusion recovery delay parameter and occlusion amplitude gradient value extracted from each data segment to form a set of light intensity occlusion feature parameters for the data channel. Arrange the light intensity occlusion feature parameter sets of all channels in chronological order to construct a dynamic detection sequence of light intensity changes.

4. The method for rapid laser thread breakage identification in the textile process according to claim 3, characterized in that: The occlusion recovery delay parameter and occlusion amplitude gradient value extracted from each data segment are recorded separately to form a set of light intensity occlusion feature parameters for the data channel, including the following steps: In the dynamic detection sequence corresponding to each detection channel, each divided data segment is traversed sequentially, and the occlusion recovery delay parameter and occlusion amplitude gradient value are extracted in each data segment. For each data segment, when a complete occlusion event is successfully identified, the occlusion recovery delay parameter corresponding to the occlusion event is stored as a time length value and bound to the data segment start timestamp as the first field of a record unit. Record the occlusion magnitude gradient value as the second field of the recording unit; The first field and the second field are combined to form a binary feature record unit, which is then associated with the data segment number and the detection channel identification information to form a set of light intensity occlusion feature parameters for the detection channel.

5. The method for rapid laser thread breakage identification in the textile process according to claim 4, characterized in that: Collect the current yarn material identification information, select the corresponding type of material response parameter group from the preset spectral response template, and establish the correspondence between yarn material and light intensity behavior characteristics, including the following steps: The yarn material identifier can be entered from the user interface or automatically read from the loom control system. Search the preset template library for the laser reflection parameters, light absorption parameters, and historical occlusion response curve parameters corresponding to the material; The found material parameters are calibrated with the current channel optical behavior feature matrix to generate channel material adaptation relationships.

6. The method for rapid laser thread breakage identification in the textile process according to claim 5, characterized in that: Based on the correspondence between yarn material and light intensity characteristics, the threshold group for determining yarn breakage in each detection channel is determined, including the following steps: The optical absorption intensity of the material is used as the adjustment benchmark for the threshold of shading abrupt change intensity; Set the upper limit of the distribution of occlusion recovery time under the same material in history as the critical time for light intensity recovery; The maximum stable amplitude of the occlusion event is used as the jump tolerance parameter. The occlusion sudden change intensity threshold, the light intensity recovery critical time, and the jump tolerance parameter together constitute the disconnection judgment threshold group of the detection channel.

7. The method for rapid laser thread breakage identification in the textile process according to claim 6, characterized in that: The current detection sequence is compared with the disconnection judgment threshold group to determine whether the light intensity meets the combined characteristics of continuous occlusion and abrupt change mode. If it does, a disconnection event is determined to have occurred in the corresponding detection channel, including the following steps: The magnitude of occlusion changes in the detected sequence is compared with the mutation intensity threshold to determine whether the mutation intensity threshold is exceeded. Determine if the duration of occlusion exceeds the critical recovery delay time; If the occlusion range exceeds the occlusion range threshold and the recovery time is greater than the recovery time threshold, then the disconnection mode characteristics are met, and a disconnection event marker is generated.

8. The method for rapid laser thread breakage identification in the textile process according to claim 7, characterized in that: The event marker for the disconnection includes the disconnection channel number, the timestamp of the disconnection, and the start and end index values ​​of the light intensity change segment before and after the disconnection in the detection sequence.

9. A method for rapid laser thread breakage identification in a textile process according to claim 8, characterized in that: The detection channel that detects a broken line collects light intensity changes within a fixed time period before and after the detection. Based on the rate of change of the occlusion edge and the stability of the fluctuation during the change segment, the occlusion abruptness threshold and jump tolerance parameters in the material response parameter group are updated, including the following steps: Extract light intensity variation data of the broken channel within a symmetrical time interval before and after the breakage; The slope of the occlusion edge in the data segment is statistically summarized, and its maximum fluctuation on the time axis is measured. The light intensity variation data and the maximum fluctuation degree are used to adjust the original abrupt change intensity threshold and jump tolerance parameters, and then updated with the original parameters using a weighted method.

10. A method for rapid laser thread breakage identification in a textile process according to claim 9, characterized in that: In the weighting method used during the update process, the historical disconnection frequency of the channel is positively correlated with the parameter update weight.

Images