Full cycle calibration method for single frequency terahertz coherent phase thickness measurement system

Abstract

The present application belongs to the technical field of terahertz detection, and particularly relates to a full-cycle calibration method of a single-frequency terahertz coherent phase thickness measurement system. The present application determines a detection reference point according to the standard parameters of a target detection object on a calibration curve, so that the same calibration process can support the reference configuration corresponding to different detection specifications, avoid repeated execution of the complete calibration process when the product specifications change, improve the flexibility of equipment use, set a review mechanism, periodically verify the effective state of the detection call data by using the standard sample, and according to the review result, select to keep the current data, adjust the compensation parameters or re-execute the full-cycle calibration, so that the equipment can adapt to the response changes generated in the long-term operation process. Through the association management among the standard sample, the calibration parameter sequence, the periodic calibration acquisition data, the effective calibration data and the detection call data, the equipment factory calibration, on-site installation, periodic maintenance and quality tracking management are facilitated.

Description

Technical Field

[0001] This invention belongs to the field of terahertz detection technology, specifically relating to a full-cycle calibration method for a single-frequency terahertz coherent phase thickness measurement system. Background Technology

[0002] Terahertz testing technology is a non-contact testing technique that utilizes the interaction between terahertz electromagnetic waves and the material under test to detect material thickness, defects, and structural parameters. Due to the strong penetrating power, high temporal resolution, and good adaptability to non-metallic materials, terahertz waves have been widely used in recent years for thickness measurement in precision manufacturing fields such as thin film materials, composite materials, coating materials, and lithium battery separators. Among these, the single-frequency terahertz coherent phase thickness measurement system transmits a fixed-frequency terahertz signal and uses a coherent receiving method to acquire the phase change of the returned signal. Based on the correspondence between the phase change and the material thickness, high-precision thickness measurement is achieved. This system features fast detection speed, no contact with the object under test, and suitability for online inspection.

[0003] In practical applications, single-frequency terahertz coherent phase thickness measurement systems typically require calibration to establish the correspondence between the coherent response and material thickness parameters. This process eliminates the influence of system component differences, environmental variations, and signal transmission path changes on the detection results. Currently, standard samples are commonly used for discrete-point calibration of the thickness measurement system. This involves selecting several standard samples with known thicknesses, collecting coherent response data under different thickness conditions, establishing the conversion relationship between thickness parameters and coherent response, and then using this conversion relationship to calculate the thickness of the actual object being measured.

[0004] Existing single-frequency terahertz coherent phase thickness measurement systems typically employ a one-time calibration method, which cannot adapt to response drift that occurs during long-term operation of the equipment. The calibration process has a limited acquisition range, making it difficult to cover different detection parameter ranges, resulting in reduced detection accuracy. Furthermore, the lack of an effective screening and periodic verification mechanism for calibration data quality makes the system susceptible to abnormal data and changes in system status, leading to insufficient reliability and long-term stability of the calibration results. Summary of the Invention

[0005] The purpose of this invention is to provide a full-cycle calibration method for a single-frequency terahertz coherent phase thickness measurement system. By establishing a full-cycle calibration mechanism, the calibration parameters of the single-frequency terahertz coherent phase thickness measurement system can be dynamically maintained, thereby improving the accuracy, stability, and long-term reliability of thickness detection.

[0006] The specific technical solution adopted by this invention is as follows: A full-cycle calibration method for a single-frequency terahertz coherent phase thickness measurement system includes: Acquire the calibration status information of the single-frequency terahertz coherent phase thickness measurement system, perform benchmark calibration on the single-frequency terahertz coherent phase thickness measurement system, and establish the initial measurement benchmark corresponding to the calibration process. Configure standard samples for calibrating the single-frequency terahertz coherent phase thickness measurement system, and determine the calibration parameter sequence based on the standard parameters corresponding to the standard samples; The standard sample is periodically scanned within the effective calibration range according to the calibration parameter sequence to obtain coherent response data and standard parameter data corresponding to different calibration positions, forming periodic calibration acquisition data; The periodic calibration data is processed for data quality, and valid calibration data that meets the preset quality conditions is selected. A calibration curve is generated based on the effective calibration data. According to the standard parameters of the target detection object, a detection reference point is determined on the calibration curve, and the reference response data corresponding to the detection reference point is obtained. Based on the calibration curve, the detection reference point, and the corresponding reference response data, generate the detection call data; In the actual testing process, the corresponding testing call data is called according to the current testing object, and the offset information between the coherent response data and the testing reference point is obtained to determine the testing parameters of the tested object; When the preset verification conditions are met, the detection call data is verified using standard samples, and the processing method corresponding to the detection call data is determined based on the verification results: retain for use, compensate for adjustment, or re-execute full-cycle calibration.

[0007] In a preferred embodiment, the calibration status information of the single-frequency terahertz coherent phase thickness measurement system is obtained, and the single-frequency terahertz coherent phase thickness measurement system is calibrated to establish an initial measurement benchmark corresponding to the calibration process, including: Acquire the calibration status information corresponding to the single-frequency terahertz coherent phase thickness measurement system, wherein the calibration status information includes terahertz transmission status information, reception status information, signal acquisition status information and phase demodulation status information; Determine whether the single-frequency terahertz coherent phase thickness measurement system meets the preset calibration conditions based on the equipment status information. When the single-frequency terahertz coherent phase thickness measurement system meets the preset calibration conditions, the terahertz transmitting module, coherent receiving module, signal acquisition module and phase demodulation module are initially calibrated. Obtain the output response data after the initial calibration is completed, and determine the initial measurement benchmark corresponding to the calibration process based on the output response data.

[0008] In a preferred embodiment, standard samples are configured for calibrating the single-frequency terahertz coherent phase thickness measurement system, and a calibration parameter sequence is determined based on the standard parameters corresponding to the standard samples, including: Obtain the detection requirements corresponding to the object to be tested, and configure standard samples for calibrating the single-frequency terahertz coherent phase thickness measurement system; Obtain the standard parameter information corresponding to the standard sample, and determine the parameter coverage range corresponding to the standard sample based on the standard parameter information; The calibration acquisition range is determined based on the parameter coverage range, and the calibration acquisition range is divided according to the preset scanning rules to obtain the calibration parameters corresponding to multiple calibration positions. The execution order corresponding to each calibration parameter is determined based on multiple calibration parameters and preset scanning rules, and a calibration parameter sequence is generated.

[0009] In a preferred embodiment, the standard sample is periodically scanned within the effective calibration interval according to the calibration parameter sequence to acquire coherent response data and standard parameter data corresponding to different calibration positions, forming periodic calibration acquisition data, including: Determine the scanning path and scanning nodes corresponding to the standard sample based on the calibration parameter sequence; The standard sample is controlled to move sequentially to each scanning node according to the scanning path within the effective calibration range to obtain the corresponding calibration position; When the standard sample is at each calibration position, acquire the coherent response data output by the single-frequency terahertz coherent phase thickness measurement system. The coherent response data includes phase response data and potential response data, and read the standard parameter data corresponding to the calibration position. Based on the coherent response data and standard parameter data corresponding to each calibration location, periodic calibration acquisition data is generated.

[0010] In a preferred embodiment, the periodic calibration data is subjected to data quality processing to filter valid calibration data that meets preset quality conditions, including: Based on the periodic calibration data, obtain the coherent response data, standard parameter data and corresponding acquisition status information for each calibration position. Based on the acquisition status information, perform data correction processing on the coherent response data and standard parameter data for each calibration position to obtain the corrected calibration data. According to the preset quality evaluation rules, the quality of the calibrated data is evaluated to obtain the data quality evaluation results for each calibration location. Obtain quality matching information between the data quality evaluation results and the preset quality conditions, and determine the calibration data that meets the preset quality conditions as valid calibration data.

[0011] In a preferred embodiment, a calibration curve is generated based on valid calibration data. According to the standard parameters of the target object, a detection reference point is determined on the calibration curve, and the reference response data corresponding to the detection reference point is obtained, including: Based on the standard parameter data and coherent response data in the effective calibration data, a mapping between calibration parameters and response parameters is established, and calibration curves are generated according to the mapping. Obtain the standard parameter information corresponding to the target detection object, and perform parameter positioning on the calibration curve based on the standard parameter information to determine the calibration position corresponding to the target detection object; The detection reference point is determined based on the calibration position corresponding to the target detection object, and the response parameters corresponding to the detection reference point are obtained. The response parameters are used as the reference response data.

[0012] In a preferred embodiment, detection call data is generated based on the calibration curve, the detection reference point, and the corresponding reference response data, including: Obtain the position of the detection reference point in the calibration curve, and determine the calibration parameter range corresponding to the detection reference point from the calibration curve; Based on the calibration parameter range, obtain the benchmark response data corresponding to the detection benchmark point, and determine the benchmark call information; Data is integrated based on calibration parameter ranges, detection benchmarks, and benchmark call information to generate detection call data for use in the online detection phase.

[0013] In a preferred embodiment, during the actual detection process, the detection parameters of the object under test are determined by calling the corresponding detection call data based on the current object being detected and obtaining the offset information between the coherent response data and the detection reference point, including: Obtain the object recognition information corresponding to the current detection object, and determine the detection call data that matches the current detection object based on the object recognition information; Obtain the detection reference point and corresponding reference response data from the detection call data, and determine the detection reference state corresponding to the current detection object; The single-frequency terahertz coherent phase thickness measurement system is controlled to detect the current object and acquire the coherent response data corresponding to the current object. The coherent response data is compared with the reference response data in the detection reference state to determine the offset information between the coherent response data and the detection reference point. Based on the offset information and detection call data, determine the detection parameters corresponding to the current detection object.

[0014] In a preferred embodiment, when preset verification conditions are met, the detection call data is verified using standard samples, and the processing method corresponding to the detection call data—whether to retain it for use, compensate for adjustments, or re-execute the full-cycle calibration—is determined based on the verification results, including: Obtain the running status information corresponding to the detection call data, and determine whether the preset review conditions have been met based on the running status information; When the running status information meets the preset review conditions, the standard sample corresponding to the detection call data is called, the standard sample is tested, and the review response data corresponding to the standard sample is obtained. Obtain the baseline response data from the detection call data, compare the review response data with the baseline response data, and determine the review offset information corresponding to the detection call data; The validity status of the detection call data is determined based on the verification offset information, and the corresponding processing method for the detection call data is determined based on the validity status.

[0015] And, the full-cycle calibration terminal of the single-frequency terahertz coherent phase thickness measurement system, including: One or more processors; A storage device on which one or more programs are stored; When one or more programs are executed by one or more processors, the one or more processors implement a full-cycle calibration method for a single-frequency terahertz coherent phase thickness measurement system.

[0016] The technical effects achieved by this invention are as follows: This invention uses a single-frequency continuous terahertz coherent phase thickness measurement method for calibration. By collecting coherent response data corresponding to different positions of a standard sample within the effective interference period, a calibration process suitable for the coherent phase detection principle is established, avoiding the calibration method using pulsed terahertz time-of-flight detection. This improves the consistency between the calibration process and the actual detection principle. By controlling the standard sample to perform periodic scanning according to the calibration parameter sequence within the effective calibration interval, the calibration process can cover the response changes within the effective detection range, obtaining complete periodic response data. Combined with data quality processing to filter effective calibration data, the impact of abnormal sampling, signal fluctuations, and other factors on the calibration results is reduced, improving the accuracy of thickness detection. By pre-generating detection call data and locking the detection reference point and reference response data corresponding to the target detection object, the online detection stage does not require a full-cycle scan. Only the corresponding detection call data needs to be called and the detection parameters need to be determined based on the real-time response offset, improving the detection response speed. This method is suitable for continuous industrial production scenarios. This invention determines the testing reference point on the calibration curve based on the standard parameters of the target testing object, enabling the same calibration process to support reference configurations corresponding to different testing specifications. This avoids repeatedly executing the complete calibration process when product specifications change, improving the flexibility of equipment use. By setting up a verification mechanism, the validity status of the test call data is periodically verified using standard samples. Based on the verification results, the current data can be kept, compensation parameters adjusted, or the full cycle calibration can be re-executed, allowing the equipment to adapt to response changes during long-term operation and reducing maintenance costs. Through the correlation management between standard samples, calibration parameter sequences, periodic calibration data, valid calibration data, and test call data, the calibration process has a clear data source and processing path, facilitating equipment factory calibration, on-site installation, periodic maintenance, and quality tracking management. Attached Figure Description

[0017] Figure 1 This is a flowchart of the method provided by the present invention. Detailed Implementation

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0019] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0020] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in a preferred embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that mutually excludes other embodiments.

[0021] Furthermore, the present invention will be described in detail with reference to the schematic diagrams. When describing the embodiments of the present invention in detail, the schematic diagrams are merely examples for ease of explanation and should not limit the scope of protection of the present invention.

[0022] Please see the appendix Figure 1 As shown, a full-cycle calibration method for a single-frequency terahertz coherent phase thickness measurement system is provided, including: S1. Obtain the calibration status information of the single-frequency terahertz coherent phase thickness measurement system, and perform benchmark calibration on the single-frequency terahertz coherent phase thickness measurement system to establish the initial measurement benchmark corresponding to the calibration process. S2. Configure standard samples for calibrating the single-frequency terahertz coherent phase thickness measurement system, and determine the calibration parameter sequence based on the standard parameters corresponding to the standard samples; S3. Control the standard sample to perform periodic scanning according to the calibration parameter sequence within the effective calibration interval, and obtain coherent response data and standard parameter data corresponding to different calibration positions to form periodic calibration acquisition data; S4. Perform data quality processing on the periodic calibration data and filter out valid calibration data that meet the preset quality conditions. S5. Generate a calibration curve based on the effective calibration data, determine the detection reference point on the calibration curve according to the standard parameters of the target detection object, and obtain the reference response data corresponding to the detection reference point. S6. Generate detection call data based on the calibration curve, detection reference point and corresponding reference response data; S7. In the actual testing process, the corresponding testing call data is called according to the current testing object, and the offset information between the coherent response data and the testing reference point is obtained to determine the testing parameters of the tested object. S8. When the preset verification conditions are met, the detection call data is verified using standard samples, and the processing method corresponding to the detection call data is determined based on the verification results: retain for use, compensate for adjustment, or re-execute full-cycle calibration.

[0023] As described in steps S1 to S8 above, during the calibration process, the operating status information of the single-frequency terahertz coherent phase thickness measurement system is acquired, and initial calibration is performed on key components such as terahertz transmission, coherent reception, signal acquisition, and phase demodulation to ensure the equipment is in a stable measurement state. An initial measurement reference is established to prevent factors such as zero-point offset of the equipment itself and signal fluctuations from affecting the calibration results. Standard samples are configured according to the thickness range, material properties, and detection accuracy requirements of the object to be tested. The calibration range and scanning sequence are determined using the standard parameters corresponding to the standard samples, ensuring that the standard samples can cover the effective interference period of the single-frequency terahertz wave in the material to be tested. By controlling the standard samples to move and scan according to a preset calibration parameter sequence, different thickness positions are brought into the measurement area. Coherent response data output by the single-frequency terahertz coherent phase thickness measurement system is acquired at each calibration position, and corresponding standard parameter data is also acquired, thus forming periodic calibration acquisition data containing thickness reference information and equipment response information. Since the actual data acquisition process may be affected by factors such as environmental changes, equipment noise, mechanical movement errors, and signal fluctuations, the periodic calibration data acquisition is processed for quality. By analyzing the acquisition status, response data stability, and standard parameter consistency, abnormal sampling points are eliminated, and data that meets the quality requirements is selected as valid calibration data. This avoids abnormal data being used to generate subsequent calibration results and improves the reliability of the calibration results. Based on the filtered valid calibration data, a correspondence between standard parameters and coherent response data is established, and a calibration curve that reflects the variation law of single-frequency terahertz coherent response is generated. According to the standard parameters corresponding to the target detection object, the corresponding detection reference point is located in the calibration curve, and the reference response data corresponding to the reference point is obtained, so that the calibration results can be associated with specific product specifications. Detection call data is generated based on the calibration curve, detection reference point and reference response data. In the actual testing process, the corresponding testing call data is called according to the specification information of the current testing object, and the real-time coherent response data of the current object is collected. By comparing the offset between the real-time coherent response data and the benchmark response data corresponding to the testing benchmark point, and combining the calibration information in the testing call data, the testing parameters of the tested object can be quickly determined without having to perform a complete cycle scan again, thus improving the efficiency of online testing. Meanwhile, to ensure the continued validity of calibration results during long-term operation, when the equipment operation reaches the preset conditions, the detection call data is reviewed using standard samples. By comparing the offset between the reviewed response data and the baseline response data, it is determined whether the current detection call data still meets the usage requirements. When the offset is within the allowable range, the current detection call data is kept in use. When there is a compensable offset, the relevant parameters of the detection baseline are adjusted. When the offset exceeds the allowable range, the full cycle calibration is re-executed, thus forming a closed-loop maintenance process of calibration, call, review and update. Calibration is performed using a single-frequency continuous terahertz coherent phase thickness measurement method. By collecting coherent response data corresponding to different positions of a standard sample within the effective interference period, a calibration process suitable for the coherent phase detection principle is established, avoiding the calibration method using pulsed terahertz time-of-flight detection. This improves the consistency between the calibration process and the actual detection principle. By controlling the standard sample to perform periodic scanning according to the calibration parameter sequence within the effective calibration interval, the calibration process can cover the response changes within the effective detection range, obtaining complete periodic response data. Combined with data quality processing, effective calibration data is filtered to reduce the impact of abnormal sampling, signal fluctuations, and other factors on the calibration results, thereby improving the accuracy of thickness detection. By pre-generating detection call data and locking the detection reference point and reference response data corresponding to the target detection object, the online detection stage does not require a full-cycle scan. It only needs to call the corresponding detection call data and determine the detection parameters based on the real-time response offset, improving the detection response speed. This method is suitable for continuous industrial production scenarios. By determining the test reference point on the calibration curve based on the standard parameters of the target test object, the same calibration process can support the reference configuration corresponding to different test specifications. This avoids repeatedly executing the complete calibration process when product specifications change, improving the flexibility of equipment use. By setting up a verification mechanism, the validity status of the test call data is verified periodically using standard samples. Based on the verification results, the system can choose to keep the current data, adjust the compensation parameters, or re-execute the full cycle calibration. This allows the equipment to adapt to response changes that occur during long-term operation, reducing maintenance costs. Through the correlation management between standard samples, calibration parameter sequences, periodic calibration data, valid calibration data, and test call data, the calibration process has a clear data source and processing path, facilitating equipment factory calibration, on-site installation, periodic maintenance, and quality tracking management.

[0024] In a preferred embodiment, the calibration status information of the single-frequency terahertz coherent phase thickness measurement system is acquired, and the single-frequency terahertz coherent phase thickness measurement system is calibrated to establish an initial measurement reference corresponding to the calibration process, including: S101. Obtain the calibration status information corresponding to the single-frequency terahertz coherent phase thickness measurement system, wherein the calibration status information includes terahertz transmission status information, reception status information, signal acquisition status information and phase demodulation status information. S102. Determine whether the single-frequency terahertz coherent phase thickness measurement system meets the preset calibration conditions based on the equipment status information. When the single-frequency terahertz coherent phase thickness measurement system meets the preset calibration conditions, the terahertz transmitting module, coherent receiving module, signal acquisition module and phase demodulation module are initially calibrated. S103. Obtain the output response data after the initial calibration is completed, and determine the initial measurement benchmark corresponding to the calibration process based on the output response data.

[0025] As described in steps S101 to S103 above, before performing the full-cycle calibration of the single-frequency terahertz coherent phase thickness measurement system, it is necessary to obtain the current operating status of the equipment to determine whether the equipment has stable calibration conditions. Since the single-frequency terahertz coherent phase thickness measurement system relies on the terahertz transmitting module to generate continuous terahertz waves, which are reflected by the sample under test and received by the coherent receiving module, then the electrical signal is acquired by the signal acquisition module, and the coherent response information is extracted by the phase demodulation module, any change in the state of any link will affect the subsequent calibration results. The operating information of each functional module is obtained through the equipment's internal state acquisition interface. Specifically, the terahertz transmission state information is obtained by reading the terahertz... The operating parameters of the transmitting module are obtained, such as the current output power, operating frequency, drive current, and continuous operating time. The control system generates control commands according to the calibration process and coordinates the terahertz transmitting module, coherent receiving module, signal acquisition module, sample moving mechanism, and data processing module to perform calibration acquisition, data processing, and detection processes. The control system sends status read commands to the terahertz transmitting module, and the transmitting module returns the current operating parameters. For example, if the detected transmitting frequency is 300 GHz, the drive current is 120 mA, and the output power is stable within the preset range, the corresponding terahertz transmitting status information is generated. The status information is received by reading the coherent receiving module. The signal reception status of the receiving module is obtained, such as the working status of the receiving detector, the received signal amplitude, the received noise level, and the stability of the optical path. This is achieved by acquiring the received signal at multiple consecutive time points after the device starts up, for example, acquiring 1000 sampling points continuously. The signal amplitude of the 1000 sampling points is then statistically analyzed to determine whether the received signal is within a stable range. Signal acquisition status information is obtained by acquiring the sampling parameters of the signal acquisition module, such as the sampling frequency, sampling resolution, sampling channel status, and acquisition buffer status. This is also achieved by reading the current configuration of the data acquisition controller, such as setting the sampling frequency to 10kHz, the sampling precision to 16-bit, and the acquisition channel to normal. If enabled, the signal acquisition status is confirmed to be normal. The phase demodulation status information is obtained by reading the output status of the phase demodulation module. For example, the continuity of the phase demodulation result, the phase output range, and the operating status of the demodulation module are obtained. By running the phase demodulation module without standard samples, the output phase values ​​within a fixed time period are collected, such as continuously acquiring 500 sets of phase data, to determine whether there are abnormal jumps between adjacent sampling points. The calibration status information obtained in the above way can comprehensively reflect the working status of each component module of the equipment, avoid calibration data distortion due to abnormal transmission power, attenuation of received signal, abnormal sampling, or unstable phase demodulation, and improve the reliability of the calibration process. After acquiring the calibration status information, the system determines whether the current single-frequency terahertz coherent phase thickness measurement system meets the preset calibration conditions based on various equipment status parameters. The preset calibration conditions are used to limit the stable operating state that the equipment must reach when entering the calibration process. The preset calibration conditions include terahertz transmission stability conditions, received signal stability conditions, acquisition status stability conditions, and phase demodulation stability conditions. Among them, the terahertz transmission stability conditions are used to determine whether the transmission module has reached a stable output state. For example, it is required that the transmission frequency fluctuation range does not exceed the set range. If the equipment is set to operate at 300 GHz, the actual output frequency change within 10 seconds of continuous detection should not exceed 0.At 1GHz, the transmission status is considered to meet the requirements. The received signal stability condition is used to determine whether the receiver can continuously acquire valid reflected signals. For example, continuously acquiring 1000 sets of received signals and comparing the signal amplitudes, if the difference between the maximum and minimum amplitudes does not exceed 5% of the average amplitude, the reception status is considered to meet the requirements. The signal acquisition stability condition is used to determine whether the acquisition module is operating according to the set parameters. For example, checking whether the sampling frequency reaches the set value of 10kHz, whether the sampling channel is normal, and whether there is data loss. If there are no data loss points during continuous acquisition, the requirements are considered to be met. Phase resolution... The stability condition is used to determine whether the demodulation module output is continuous. For example, if 500 sets of phase output data are continuously acquired, and there are no abnormal abrupt changes between adjacent data, and the phase output change is within the allowable range of the device, then the requirement is considered met. Based on the above judgment result, it is determined whether the device meets the preset calibration conditions. When all status information meets the corresponding conditions, the initial calibration stage begins. If any status information does not meet the requirements, the device is prompted to continue warming up, adjusting parameters, or re-testing until the calibration conditions are met. After the calibration conditions are met, the terahertz transmitting module, coherent receiving module, signal acquisition module, and phase demodulation module are calibrated respectively. Initial calibration is performed. When calibrating the terahertz transmitting module, the module outputs a fixed-frequency terahertz signal, and the current stable output state is recorded as the transmission reference. For example, the transmission frequency is set to 300 GHz, and output data is continuously collected for 10 seconds. The average output value under stable conditions is used as the transmission calibration value. When calibrating the coherent receiving module, a reference response is generated by inputting a fixed reflection signal or a standard reflector. The gain and zero point of the receiving channel are adjusted accordingly. For example, 100 sets of received signals are collected using fixed reflection samples, and the average value of these 100 sets of signals is used as the receiving reference value. The signal acquisition module is then calibrated. During line calibration, a known amplitude signal is input to correct the acquisition channel deviation. For example, a 1V standard voltage signal is input, and the acquisition result is checked to see if it is close to 1V. The acquisition parameters are adjusted according to the deviation. When calibrating the phase demodulation module, a zero-phase output state is obtained by fixing a reference signal. This state is used as the initial reference for phase demodulation. For example, a reference signal with no thickness change is input, and the corresponding output phase value is recorded as the phase zero point. Through the above initial calibration process, different functional modules are all in a unified reference state before the calibration begins, reducing the impact of internal equipment deviations on the subsequent full-cycle calibration process and improving the consistency of calibration results. After the device completes module calibration, it maintains a stable operating state and continuously acquires the device's output signal within a preset time. For example, after the device is running stably, it continuously acquires output response data within 5 seconds, acquiring 100 sets of data per second, for a total of 500 sets of response data. Each set of data includes the coherent response value, phase value, and signal potential value corresponding to the current time. Stability processing is performed on the acquired sets of output response data. For example, the 500 sets of data are arranged in chronological order of acquisition time, and the changes between data points at each time point are calculated. When the data change is within the allowable range, the set of data is determined as valid output response data. Based on the output response data in a stable state, the baseline response value when the device is not loaded with a standard sample or is in a reference state is determined. For example, 500 sets of potential response data are continuously acquired, where the first to the 500th sets of data are the device's output potential values. By averaging the 500 sets of data, the baseline response value is determined. The average potential value obtained by summing the data sets and dividing by 500 is used as the initial potential reference. Similarly, the average phase value is obtained by processing the 500 sets of coherent response data and used as the initial phase reference. The initial phase reference, the initial potential reference, and the corresponding equipment status information are associated and stored to form the initial measurement reference for the calibration process. For example, if the average phase value is 30° and the average potential value is 2.5V when the equipment is in a stable state, then 30° is used as the initial phase reference value and 2.5V is used as the initial potential reference value. In the subsequent standard sample scanning process, these are used to determine whether the change in the acquired response comes from the change in sample thickness rather than the equipment's own drift. By establishing the initial measurement reference, the coherent response data obtained in the subsequent periodic scanning process all have a unified reference starting point, improving the consistency of the calibration process at different times and in different batches, while reducing the calibration error caused by equipment temperature drift, zero-point drift, and electronic noise.

[0026] In a preferred embodiment, standard samples are configured for calibrating the single-frequency terahertz coherent phase thickness measurement system, and a calibration parameter sequence is determined based on the standard parameters corresponding to the standard samples, including: S201. Obtain the detection requirements corresponding to the object to be tested, and configure the standard sample for calibrating the single-frequency terahertz coherent phase thickness measurement system. S202. Obtain the standard parameter information corresponding to the standard sample, and determine the parameter coverage range corresponding to the standard sample based on the standard parameter information; S203. Determine the calibration acquisition range based on the parameter coverage range, and divide the calibration acquisition range according to the preset scanning rules to obtain the calibration parameters corresponding to multiple calibration positions; S204. Determine the execution order corresponding to each calibration parameter based on multiple calibration parameters and preset scanning rules, and generate a calibration parameter sequence.

[0027] As described in steps S201 to S204 above, when calibrating the single-frequency terahertz coherent phase thickness measurement system, it is necessary to clarify the testing requirements of the object under test in the actual production process to determine the calibration range, standard sample type, and scanning accuracy requirements. The testing requirement information corresponding to the object under test is obtained through production task information, product process parameters, or testing equipment configuration files. This testing requirement information includes at least the type of object under test, material type, target testing parameters, testing range, and allowable testing deviation. The product information corresponding to the current product batch is read from the production line management system by the control system. For example, if the current object under test is a lithium battery electrode coating, the process file corresponding to that batch is read, and the target testing parameter is the coating thickness, with a target thickness range of 80μm to 120μm and an allowable thickness deviation of ±2μm. This information is then used as the current testing requirement. Based on the testing requirement, standard samples are configured for calibrating the single-frequency terahertz coherent phase thickness measurement system. The standard sample with similar terahertz propagation characteristics is selected based on the material type, thickness range, and detection accuracy of the object under test. For example, when the object under test is a lithium battery electrode coating, a standard thin film sample with dielectric properties close to the actual coating material is selected as the calibration sample. Multiple standard samples with different thicknesses are configured according to the detection range. For example, standard thin film samples with thicknesses of 70μm, 80μm, 90μm, 100μm, 110μm, 120μm, and 130μm are configured so that the standard sample covers the target detection thickness range and its upper and lower extension range. If it is not possible to cover the complete range with multiple fixed thickness samples, a fixed thickness standard sample can be configured, and the equivalent thickness position between the sample and the detection probe can be changed by combining a precision displacement mechanism. For example, a standard sheet with a thickness of 100μm is used, and the equivalent thickness change is obtained by moving it in 1μm increments through a precision displacement mechanism, so that the calibration process can cover the thickness variation range of the object under test in actual production. After configuring the standard samples, it is necessary to obtain the standard parameter information corresponding to each standard sample to determine the effective calibration range that the standard sample can cover. The standard parameter information mainly includes the actual thickness value, material type, sample number, metrological error, and corresponding effective usage range of the standard sample. The standard parameters are obtained through the metrological testing records, manufacturing identification information, or third-party metrological data of the standard sample. For example, for the standard film sample numbered A001, multi-point measurements are performed using a precision thickness measuring device to obtain the thickness values ​​at different locations of the sample. If the thicknesses at the five locations are measured to be 99.8μm, 100.1μm, 100μm, 100.2μm, and 99.9μm, then the above multiple test results are statistically processed to determine that the nominal thickness of the standard sample is 100μm, and the corresponding measurement error range is recorded. For the standard sample group, the standard parameter information corresponding to each sample is obtained separately. For example, 70μm, 80μm, 90μm, 100μm, 110μm, and 120μm samples correspond to different thicknesses. The parameters are determined based on the minimum and maximum values ​​of the standard sample thickness parameters to establish the basic coverage range, which is then expanded according to the detection requirements of the object under test. For example, if the minimum thickness in the standard sample group is 70 μm and the maximum thickness is 130 μm, the basic coverage range is 70 μm to 130 μm. If the detection requirement of the object under test is 80 μm to 120 μm, it is determined that the coverage range of the standard sample can cover the target detection range. To ensure that the single-frequency terahertz coherent phase thickness measurement system can completely identify the coherent response period, it is also necessary to determine whether the coverage range covers the thickness variation range corresponding to an effective interference period. For example, if the effective thickness period in the material corresponding to the current terahertz wavelength is about 200 μm, it is necessary to use standard samples or displacement mechanisms to ensure that the calibration range can cover at least 200 μm of the equivalent variation range. If the fixed samples are insufficient to cover the area, the number of samples is increased or the displacement scanning range is expanded. By determining the coverage range of the standard sample parameters, the calibration acquisition can be planned for the effective detection range, avoiding the inability to completely identify the phase period due to insufficient calibration range, and improving the accuracy of the calibration results. The calibration acquisition range is determined jointly based on the detection range of the object under test, the coverage range of the standard sample, and the effective interference period corresponding to the single-frequency terahertz coherent phase thickness measurement system. For example, if the detection range of the object under test is 80μm to 120μm, the coverage range of the standard sample is 70μm to 130μm, and the corresponding effective interference period is 200μm, then 70μm to 130μm can be used as the basic calibration acquisition range to cover the target thickness area while retaining a certain boundary margin. After determining the calibration acquisition range, the range is divided according to the preset scanning rules. The preset scanning rules are used to specify the scanning step size, scanning direction, and sampling density. The scanning step size is determined based on the target detection accuracy. For example, if the target thickness detection accuracy is 1μm, then the scanning step size is set to 0.5μm to 1μm. The scanning direction can be set to gradually scan from small thickness to large thickness, or a reciprocating scanning method can be used to reduce mechanical movement errors. The sampling density is used to determine the sampling density per... The number of data acquisitions within a thickness range, for example, in the range of 70μm to 130μm, with a scanning step of 1μm, the range is divided into multiple calibration positions of 70μm, 71μm, 72μm...130μm, forming a total of 61 calibration positions. For the method of using a precision displacement mechanism, the equivalent calibration position is determined according to the movement distance of the displacement mechanism. For example, the initial position corresponds to 70μm, and a new equivalent thickness calibration position is formed every 1μm movement. Corresponding calibration parameters are generated according to the divided calibration positions. The calibration parameters include at least the calibration position number, the corresponding standard thickness value, the movement position parameter, and the acquisition sequence identifier. For example, the first calibration position corresponds to a thickness of 70μm, the second calibration position corresponds to a thickness of 71μm, and the 61st calibration position corresponds to a thickness of 130μm. In the above way, the continuous thickness variation range is converted into multiple discrete calibration positions, and corresponding coherent response data can be acquired at each determined position. After obtaining the calibration parameters corresponding to multiple calibration positions, it is necessary to determine the execution order of each calibration parameter so that the standard sample can complete a full scan along a predetermined path. The multiple calibration parameters are sorted according to preset scanning rules, based on factors including the direction of calibration position change, scanning continuity requirements, and mechanical movement stability requirements. For example, when the preset scanning rule is set to unidirectional scanning from low to high thickness, the standard thickness values ​​are sorted from smallest to largest, with 70μm, 71μm, 72μm arranged sequentially up to 130μm, forming an incremental execution order. When a reciprocating scanning rule is used, a reciprocating execution path is generated in the manner of 70μm, 130μm, 70μm. The position corresponding to each calibration parameter is then numbered... The calibration parameter sequence, standard thickness value, movement control parameters, and acquisition identifier are combined according to the execution order. For example, the generated sequence includes the first execution position at 70μm, the second execution position at 71μm, the third execution position at 72μm, ..., the 61st execution position at 130μm, and a corresponding data acquisition command is configured for each position. When calibration is performed, the standard sample is controlled to move to the corresponding position in sequence according to the calibration parameter sequence, and coherent response data is acquired at each position. By generating a calibration parameter sequence with a clear execution order, the problem of data loss or discontinuous scanning path caused by random scanning can be avoided, so that the entire calibration process can completely cover the effective calibration range and improve the continuity and reliability of periodic calibration data.

[0028] In a preferred embodiment, the standard sample is controlled to perform periodic scanning according to the calibration parameter sequence within the effective calibration interval, acquiring coherent response data and standard parameter data corresponding to different calibration positions, forming periodic calibration acquisition data, including: S301. Determine the scanning path and scanning nodes corresponding to the standard sample based on the calibration parameter sequence; S302. Control the standard sample to move sequentially to each scanning node according to the scanning path within the effective calibration interval to obtain the corresponding calibration position; S303. When the standard sample is at each calibration position, acquire the coherent response data output by the single-frequency terahertz coherent phase thickness measurement system, wherein the coherent response data includes phase response data and potential response data, and read the standard parameter data corresponding to the calibration position. S304. Generate periodic calibration acquisition data based on the coherent response data and standard parameter data corresponding to each calibration position.

[0029] As described in steps S301 to S304 above, multiple calibration parameters are read from the calibration parameter sequence. Each calibration parameter corresponds to a target calibration state of the standard sample. The calibration parameters can be the equivalent thickness value, displacement position value, or thickness change amount formed by adjustment by a precision displacement mechanism of the standard sample. For example, when calibrating a lithium electrode film with a thickness range of 50μm to 150μm, based on the actual thickness coverage of the standard sample and the effective interference period corresponding to the single-frequency terahertz wave, the calibration parameter sequence is set to 50μm, 55μm, 60μm, 65μm, 70μm, 75μm, 80μm, 85μm, 90μm, 95μm, 100μm, 105μm, 110μm, 115μm, 120μm, 125μm, 130μm, 135μm, 140μm, 145μm, 145μm, 15μm, 120μm, 125μm, 130μm, 135μm, 140μm, 145μm, 1 ... There are 21 calibration parameter points, including μm and 150μm. The scanning path is determined based on the relationship between the calibration parameters. When a fixed standard sample is used in conjunction with a precision displacement mechanism for calibration, the linear displacement path is determined according to the direction of thickness change. When multiple standard samples of different thicknesses are used for calibration, the sample switching path is determined according to the order of the standard sample numbers. The position corresponding to each calibration parameter is used as a scanning node. For example, 50μm corresponds to the first scanning node, 55μm corresponds to the second scanning node, and so on, forming multiple scanning nodes covering the entire effective calibration range. This ensures that the scanning path is consistent with the calibration parameter sequence, avoids discontinuity in the thickness change process due to chaotic scanning order, and ensures that the coherent response data acquired later can fully reflect the variation law within the single-frequency terahertz interference period, thereby improving the continuity and reliability of the calibration data. The control system reads the scanning path information and generates motion control commands based on the position parameters corresponding to the scanning nodes. It then controls the displacement of the standard sample via a drive motor, linear guide, piezoelectric displacement mechanism, or precision moving platform. For example, when the first scanning node corresponds to a standard sample thickness of 50μm, the precision displacement platform is moved to its initial position, placing the standard sample at the 50μm detection position. The platform is then moved sequentially to subsequent nodes at 55μm, 60μm, and 65μm in 5μm increments. During each movement, the actual position is acquired via a displacement sensor or motion control feedback signal, and compared with the target scanning node. When the deviation between the actual position and the target position is less than a preset allowable error, for example, the displacement... If the error is less than 0.5μm, the standard sample is considered to have reached the corresponding scanning node. If the deviation exceeds the allowable range, the motion mechanism is controlled to perform position compensation adjustment until the target position is reached. After the positioning of each scanning node is completed, the current spatial position of the standard sample, the corresponding thickness parameter, and the node number are associated to form the corresponding calibration position data. For example, if the current standard sample is located at the 10th scanning node and the corresponding thickness parameter is 95μm, then the node number 10, the calibration parameter 95μm, and the actual displacement position X10 are recorded as the calibration position corresponding to this position. This can ensure that the standard sample moves along the predetermined trajectory throughout the entire cycle of scanning, improve the positioning accuracy of the standard sample, and reduce the impact of mechanical movement error on the subsequent coherent response acquisition results. Once the standard sample has moved to a calibrated position, it waits for a stable sampling state, for example, 100ms to 500ms, to allow the output power of the terahertz transmitting module, the response of the receiving module, and the phase demodulation module to stabilize before initiating the acquisition process. The terahertz transmitting module emits continuous terahertz waves towards the standard sample, which are reflected from the upper and lower surfaces of the sample to form a coherent superposition signal. The receiving module acquires the reflected signal, and the phase demodulation module obtains the corresponding coherent response data. The coherent response data includes phase response data and potential response data. For example, at a scanning node with a standard sample thickness of 100μm, a coherent phase value of 135 degrees is acquired. If the simulated potential value is 2.8V, then this set of data is used as the coherent response data corresponding to the calibration position. At the same time, the standard parameter data corresponding to this position is read. The standard parameter data comes from the metrological information of the standard sample or the feedback information of the precision displacement mechanism. For example, if the actual thickness of the standard sample is determined to be 100μm after metrological measurement, then 100μm is used as the standard parameter data corresponding to this position. In order to improve the reliability of the data, multiple sets of response data can be repeatedly collected at each calibration position, for example, five consecutive collections. The results of the five collections are saved for quality screening, which can obtain the actual system response corresponding to different thickness states covering the entire effective interference period. The coherent response data and standard parameter data corresponding to each calibration location are read, and data matching is performed according to the calibration location number. For example, the first calibration location corresponds to a standard parameter of 50 μm, and the coherent response data includes a phase value of 80° and a potential value of 1.5V; the second calibration location corresponds to a standard parameter of 55 μm, and the coherent response data includes a phase value of 90° and a potential value of 1.7V. The data from all scanning nodes are combined sequentially, and corresponding auxiliary information is added to each group of data, including the calibration location number, acquisition time, ambient temperature, equipment status information, and resampling sequence number, thus forming complete periodic calibration acquisition data. For example, a group of periodic calibration acquisition data may include the node number. 1. Standard thickness 50μm, coherent response 80°, potential response 1.5V, ambient temperature 25℃, acquisition time 10:00:01. Another set of data includes node number 2, standard thickness 55μm, coherent response 90°, potential response 1.7V, ambient temperature 25℃, acquisition time 10:00:05. After all the scanning node data are integrated, a periodic calibration acquisition data covering the effective calibration range of 50μm to 150μm is formed. The discrete single-point detection results are converted into a dataset with continuous scanning characteristics. At the same time, it can ensure that the calibration process covers the single-frequency terahertz coherent response period and avoid phase period identification errors caused by insufficient sampling range.

[0030] In a preferred embodiment, the periodic calibration data is subjected to data quality processing to filter valid calibration data that meets preset quality conditions, including: S401. Obtain the coherent response data, standard parameter data and corresponding acquisition status information corresponding to each calibration position based on the periodic calibration acquisition data. Based on the acquisition status information, perform data correction processing on the coherent response data and standard parameter data corresponding to each calibration position to obtain the corrected calibration acquisition data. S402. According to the preset quality evaluation rules, the quality evaluation of the calibrated data is carried out to obtain the data quality evaluation results corresponding to each calibration position. S403. Obtain the quality matching information between the data quality evaluation results and the preset quality conditions, and determine the calibration data that meets the preset quality conditions as valid calibration data.

[0031] As described in steps S401 to S403 above, the calibration position number in the periodic calibration acquisition data is read, a data index is established based on the calibration position number, and the corresponding data content is extracted from each index position. The coherent response data originates from the detection output acquired by the single-frequency terahertz coherent phase thickness measurement system at the corresponding calibration position, including coherent response values, simulated potential response values, and sampling time information. The standard parameter data originates from the metrological information of the standard sample or feedback information from the precision displacement mechanism, including the actual thickness value, displacement, or equivalent thickness change value corresponding to the standard sample. The acquisition status information originates from the internal status record of the equipment during the calibration acquisition process, including the terahertz transmission power status, The system collects data on received signal strength, signal sampling status, phase demodulation status, ambient temperature, and equipment operating time. For example, at the 15th calibration position, the periodic calibration data shows a standard parameter of 100μm, a coherent response of 135°, and a potential response of 2.8V. Simultaneously, it records that the terahertz transmit power is 99% of the rated output, the received signal amplitude is within the normal range, the ambient temperature is 26℃, and the phase demodulation module is functioning normally. When a change in ambient temperature is detected causing a response shift, temperature compensation is applied to the coherent response data based on the temperature change range recorded during the calibration process. For instance, if the initial calibration temperature is 25℃ and the current acquisition temperature is 27℃... Based on the temperature offset data collected under multiple temperature conditions, the coherent response value is corrected. When fluctuations in transmission power are detected, the simulated potential response is normalized based on the difference between the actual transmission power and the standard output value. For example, if the standard transmission power corresponds to a potential response of 2.8V, and the current transmission power decreases by 1%, the potential response is compensated according to the power change ratio. When random noise is detected in repeated sampling, multiple sampling data from the same calibration location are averaged. For example, if five consecutive potential response values ​​at the same thickness location are 2.79V, 2.81V, 2.8V, 2.82V, and 2.78V, then the average is calculated. The average of five data points is calculated, and 2.8V is taken as the final response value at that location. After calibration, the calibrated coherent response data, standard parameter data, and corresponding acquisition status information are recombined to form the calibrated acquisition data. For example, the original data of 100μm thickness, 135 degrees phase, and 2.8V is compensated for temperature and power to obtain 100μm thickness, corrected phase of 134.6°, and corrected potential of 2.82V. This data is used as the calibrated acquisition data, which can reduce the impact of equipment drift, environmental changes, and acquisition anomalies on the calibration data, improve the reliability of the data used to generate the calibration curve, and prevent erroneous data from being solidified as the basis for long-term use of the equipment. Pre-defined quality evaluation rules are established based on the detection characteristics of the single-frequency terahertz coherent phase thickness measurement system. These rules include response continuity evaluation rules, resampling consistency evaluation rules, signal stability evaluation rules, and parameter matching evaluation rules. The response continuity evaluation rule is used to determine whether the coherent response changes between adjacent calibration positions conform to a continuous trend. For example, when the standard parameter increases from 50 μm to 55 μm, the coherent response increases from 80° to 90°. If an adjacent position suddenly jumps from 80° to 200°, it is considered an anomaly at that position. The resampling consistency evaluation rule is used to determine whether the results of multiple acquisitions at the same calibration position are stable. For example, if five phase acquisitions at the same location yield 100°, 101°, 99°, 100°, and 102° respectively, the repeatability is considered good. Conversely, if the results are 100°, 150°, 80°, 170°, and 90° respectively, the repeatability is considered poor. Signal stability evaluation rules are used to determine whether the device output is stable during acquisition. For example, requirements include a terahertz transmit power fluctuation range of less than 2% and a received signal strength change of less than 5%. Parameter matching evaluation rules are used to determine whether standard parameter data corresponds to the calibration position. For example, if the standard sample thickness is recorded as 100μm, but the displacement mechanism feedback position corresponds to 95μm, then a parameter discrepancy exists at that position. Typically, the data corresponding to each calibration position is evaluated according to the aforementioned quality evaluation rules. For example, when evaluating the data at the 15th calibration position, it is determined whether its coherent response is within the range of variation of adjacent nodes. It is found that the phase of the previous node is 130°, the current node is 134.6°, and the next node is 139°, which meets the requirement of continuous variation. Next, it is determined whether the deviation of the results of 5 repeated samplings is less than the preset range. For example, the difference between the maximum and minimum values ​​is 1.5°, which is less than the allowable deviation of 3°, thus meeting the repeatability requirement. Then, it is determined whether the acquisition status information is normal. It is found that the transmission power fluctuation is 0.8% and the temperature change is 0.5℃, both of which meet the requirements. Finally, the standard thickness is determined. To verify if the calibration curve matches the actual location, the standard thickness is confirmed to be 100μm, meeting the matching requirements. Therefore, this location is evaluated as qualified data. Data quality evaluation results corresponding to this calibration location can be generated based on multiple evaluation results. For example, three status indicators can be used: qualified, pending review, and unqualified. Data that meets all evaluation conditions is marked as qualified, data with only minor anomalies that can be corrected is marked as pending review, and data with serious anomalies is marked as unqualified. Abnormal collection points can be identified in advance before the calibration curve is generated, avoiding mechanical movement errors, signal interference, or abnormal equipment status from causing erroneous calibration data to enter the subsequent calculation process, thereby improving the accuracy and stability of the calibration results. Preset quality conditions are obtained, derived from the calibration accuracy requirements of the single-frequency terahertz coherent phase thickness measurement system and industrial testing needs. These conditions include requirements such as continuous phase change between adjacent calibration positions, resampling error less than 3 degrees, equipment output stability meeting transmit power fluctuation requirements of less than 2%, ambient temperature variation range of less than 5°C, and standard parameter error less than 1 μm from actual positioning. The data quality evaluation results for each calibration position are then matched against these quality conditions. For example, if the data quality evaluation results for the 15th calibration position include satisfactory response continuity, satisfactory resampling, satisfactory equipment status, and satisfactory parameter matching, then it fully matches the preset quality conditions, and the calibration data collected at that position is determined to meet the conditions. However, if the data quality evaluation results for the 20th calibration position show abnormal response continuity (e.g., a sudden change in phase value from 150° to 230°, inconsistent with the trend of adjacent nodes), then the data is determined to be unsatisfactory. Data that does not meet the quality conditions is not considered valid calibration data. When some data does not meet the quality conditions, different processing can be performed according to the type of anomaly. For example, for a single abnormal sampling point, data at the corresponding calibration position can be re-acquired. For multiple consecutive abnormal positions, the scanning of that area is re-executed. If there are anomalies in the entire cycle data, the full cycle scan is re-executed. After screening, all data that meet the preset quality conditions are rearranged according to the calibration position order to form valid calibration data. For example, if the original cycle calibration acquisition data contains 21 calibration positions, of which 19 positions meet the quality conditions and 2 positions are removed due to excessive repeated sampling errors, then 19 sets of data are retained as valid calibration data. This ensures that the data entering the subsequent calibration curve generation stage has high reliability, so that the final detection call data can accurately reflect the actual response characteristics of the single-frequency terahertz coherent phase thickness measurement system, improving the measurement accuracy and stability of the equipment during long-term operation.

[0032] In a preferred embodiment, a calibration curve is generated based on effective calibration data. A detection reference point is determined on the calibration curve according to the standard parameters of the target object being detected, and the reference response data corresponding to the detection reference point is obtained, including: S501. Based on the standard parameter data and coherent response data in the effective calibration data, establish the corresponding mapping between calibration parameters and response parameters, and generate calibration curves according to the corresponding mapping; S502. Obtain the standard parameter information corresponding to the target detection object, and perform parameter positioning on the calibration curve based on the standard parameter information to determine the calibration position corresponding to the target detection object. S503. Determine the detection reference point based on the calibration position corresponding to the target detection object, and obtain the response parameters corresponding to the detection reference point, and use the response parameters as the reference response data.

[0033] As described in steps S501 to S503 above, standard parameter data and coherent response data corresponding to each calibration position are read from the valid calibration data. The standard parameter data originates from the measurement results of the standard sample or the feedback results of the precision displacement mechanism. For example, a calibration position corresponds to a standard sample thickness of 50μm, 55μm, 60μm, 65μm, 70μm, etc. The coherent response data originates from the data collected by the single-frequency terahertz coherent phase thickness measurement system when detecting the corresponding standard sample, including coherent response values ​​and simulated potential response values. For example, the valid calibration data includes a standard parameter of 50μm corresponding to a coherent response of 80° and a simulated potential response of 1.5V, and a standard parameter of 60μm corresponding to... The coherent response is 100°, and the simulated potential response is 1.9V. The standard parameter 70μm corresponds to a coherent response of 120° and a simulated potential response of 2.3V. Each standard parameter data point is associated with its corresponding coherent response data, binding standard parameters and response parameters at the same calibration location to form a set of calibration parameter-response parameter correspondence data. For example, 50μm, 80°, 1.5V; 60μm, 100°, 1.9V; 70μm, 120°, 2.3V. After completing the association of all valid calibration data, the corresponding data are sorted according to the order of standard parameter changes, arranging the standard parameters from smallest to largest while maintaining the corresponding response parameters in a synchronized arrangement, thus forming continuous calibration data. According to the sequence, a calibration curve is generated based on the calibration data sequence. This can be achieved using discrete data connection, piecewise fitting, or lookup table methods. When using discrete connection, each standard parameter point is used as the horizontal position, and the corresponding response parameter is used as the vertical position. The data points are connected in ascending order of the standard parameters to form a curve showing the response change corresponding to the change in the standard parameter. For example, 50μm corresponds to 1.5V, 60μm corresponds to 1.9V, and 70μm corresponds to 2.3V, thus forming a curve showing the gradual change in potential response from 50μm to 70μm. When using piecewise fitting, the adjacent intervals are made continuous based on the changing trend between multiple adjacent calibration points. For example... A variation trend is used between 50μm and 60μm, and another variation trend is used between 60μm and 70μm, so that any detection response can locate the corresponding thickness position. When the lookup table method is used, the correspondence between standard parameters and response parameters is directly saved. For example, data such as phase 80° and potential 1.5V corresponding to a thickness of 50μm and phase 90° and potential 1.7V corresponding to a thickness of 55μm are saved. This converts the effective calibration data obtained from discrete acquisition into calibration curves that can be called up for online detection, and establishes a stable correspondence between the coherent response change in the single-frequency terahertz system and the actual thickness change, avoiding complex recalculation in the online detection stage and improving detection speed and detection result stability. Obtain the standard parameter information corresponding to the target inspection object. This standard parameter information can come from production process settings, product specification databases, manufacturing execution systems (MES), or manually input information. For example, in the production process of lithium electrode sheets, if the current product model is A, the design thickness parameters corresponding to this model can be read from the production system to obtain a standard thickness of 100μm for the target inspection object, with an allowable thickness fluctuation range of ±2μm. Alternatively, in the thin film production process, the current product specification can be read from the product order information to determine a target inspection thickness of 80μm. It is then determined whether the target standard parameter is within the coverage range of the calibration curve. For example, if the calibration curve covers 50μm to 150μm, and the standard thickness of the target inspection object is 100μm, then the parameter is determined to be within the valid range. The data position corresponding to 100μm is then found on the calibration curve. If a calibration point corresponding to 100μm exists on the calibration curve, that point is directly used to determine the target thickness. The target detection position is determined by interpolation based on adjacent calibration points if no perfect matching point exists. For example, in the calibration curve, 90μm corresponds to a response parameter of 130°, and 110μm corresponds to a response parameter of 170°. The target standard thickness is 100μm. Based on the trend between 90μm and 110μm, the response position corresponding to 100μm is determined to be approximately 150°. The curve position corresponding to this response position is determined as the calibration position corresponding to the target detection object. For example, for a 100μm thick product, the 100μm node in the calibration curve is determined as the target calibration position. This position corresponds to a coherent response of 150° and a potential response of 2.5V. This allows different product specifications to quickly determine the corresponding detection position based on the same calibration, realizing the reuse of calibration data between different product specifications. It avoids re-performing the complete calibration process every time the product specification is switched, improving equipment adaptability and production switchover efficiency. Based on the calibration location of the target object, the data point corresponding to that location is read from the calibration curve or calculated, and this location is defined as the detection reference point. The detection reference point includes at least the standard parameter value corresponding to the target object and the corresponding response parameter value. For example, if the standard thickness of the target object is 100μm, and the location corresponding to 100μm is found on the calibration curve, then the standard thickness of 100μm, coherent response of 150°, and potential response of 2.5V are determined as the detection reference point information. The coherent response value corresponding to the detection reference point is read from the data set corresponding to the calibration curve. If directly corresponding data exists in the calibration curve, it is read directly; for example, 100μm corresponds to a coherent response of 150° and a potential response of 2.5V. If no directly corresponding data exists, it is calculated based on adjacent calibration points; for example, 95μm corresponds to a coherent response of 150° and a potential response of 2.5V. A response of 140° corresponds to a coherent response of 160° for 105μm. Based on the trend of these two points, a coherent response of 150° is determined for 100μm. The obtained response parameters are saved as reference response data. For example, for a 100μm thick product, the reference response data is saved as a reference phase of 150°, a reference potential of 2.5V, and a corresponding standard thickness of 100μm. When the actual coherent response data obtained from the test deviates from this reference response data, the actual thickness change is determined based on the direction and degree of deviation. The complete calibration curve is converted into a test reference for specific product specifications. It is not necessary to retrace the entire calibration range in each test. Only the corresponding test reference needs to be called to complete the rapid thickness judgment, which improves the efficiency of online testing and reduces the risk of misjudgment caused by periodic phase changes, thereby improving the stability of continuous industrial testing.

[0034] In a preferred embodiment, detection call data is generated based on the calibration curve, detection reference point, and corresponding reference response data, including: S601. Obtain the position of the detection reference point in the calibration curve, and determine the calibration parameter range corresponding to the detection reference point from the calibration curve; S602. Obtain the benchmark response data corresponding to the detection benchmark point based on the calibration parameter range, and determine the benchmark call information; S603. Integrate the data based on the calibration parameter range, detection reference point and reference call information to generate detection call data for use in the online detection stage.

[0035] As described in steps S601 to S603 above, the detection reference point information is read. This information originates from a defined detection reference point and includes the standard parameter values ​​of the target object and the corresponding reference response data. For example, if the target object is a lithium electrode sheet with a standard thickness of 100 μm, the corresponding coherent response of the detection reference point is 150°, and the simulated potential response is 2.5V. All calibration parameter nodes in the calibration curve are also read. For example, if the calibration curve covers a range of 50 μm to 150 μm, the corresponding nodes include 50 μm, 60 μm, 70 μm, 80 μm, and 90 μm. For thicknesses of 0μm, 100μm, 110μm, 120μm, 130μm, 140μm, and 150μm, the standard thickness of 100μm corresponding to the detection reference point is matched with the above calibration parameter nodes. When a node corresponding to 100μm exists in the calibration curve, the 100μm node is directly determined as the detection reference point position. When no perfectly corresponding node exists, the corresponding position is determined based on the changing trend between adjacent nodes. For example, in the calibration curve, 90μm corresponds to a coherent response of 130°, and 110μm corresponds to a coherent response of 170°, while the target detection thickness is... If the thickness is 100μm, then the location between the 90μm and 110μm nodes is determined as the detection reference point. After determining the detection reference point, the corresponding interval is selected based on the allowable thickness variation range of the target object and the coverage range of the calibration curve. For example, if the standard thickness of the target product is 100μm and the allowable deviation range in production is ±3μm, then the actual detection focus range is 97μm to 103μm. To ensure that deviation judgment can be completed during the detection process, a detection margin can be added to this range, for example, by extending it by 5μm to both sides, ultimately determining the range to be 92μm to 103μm. 08μm is used as the corresponding calibration parameter range. Data nodes in the range of 92μm to 108μm are extracted from the calibration curve. For example, the response data corresponding to the three calibration positions of 95μm, 100μm, and 105μm are used as the local calibration range corresponding to the detection benchmark point. This can convert the large range of data obtained from the full cycle calibration into a local calling range for the current product specification, reduce the data search range in the online detection process, and avoid calling data from non-target cycle positions due to the periodicity of the single-frequency terahertz coherent response, thereby improving detection accuracy. The calibration parameter range is determined. For example, if the calibration parameter range for the current product is determined to be 92μm to 108μm, the location corresponding to the detection reference point is found within this range, and the reference response data corresponding to that location is read. The reference response data comes from the coherent response parameters corresponding to the detection reference point in the calibration curve, including the reference coherent response value, the reference potential response value, and the corresponding standard parameter value. For example, at the 100μm detection reference point, the calibration curve records a coherent response of 150° and a potential response of 2.5V. Therefore, 150° and 2.5V are used as the reference response data corresponding to this detection reference point. Simultaneously, to ensure that online detection can perform thickness conversion based on real-time detection results, it is also necessary to obtain response change information near the reference point from the calibration parameter range. For example, within the 92μm to 108μm range, data corresponding to a coherent response of 140° at 95μm, 150° at 100μm, and 160° at 105μm are read. The thickness is determined using the above data. The correspondence between degree change and response change is established. Based on the calibration parameter range, detection reference point, and reference response data, the reference call information is determined. The reference call information describes the data content that needs to be called during the online testing phase, including the current product specification number, the location of the detection reference point, the applicable calibration parameter range, the reference response value, and the response change range. For example, for a 100μm electrode product, the reference call information includes the product specification number A001, the detection reference thickness of 100μm, the call range of 92μm to 108μm, the reference phase of 150°, the reference potential of 2.5V, and the callable response range of 140° to 160°. The information comes from the calibration curve and the detection reference point data, and is generated through data filtering and association. Key detection information in the calibration curve can be extracted and organized, so that during online testing, it is not necessary to traverse the entire calibration curve. Only the corresponding reference call information needs to be called to complete the test, which improves the system response speed and reduces data matching errors caused by excessively large call ranges. The system reads a defined calibration parameter range, such as 92μm to 108μm, and a defined detection reference point information, such as a standard thickness of 100μm, a reference phase of 150°, and a reference potential of 2.5V. Simultaneously, it reads reference call information, such as product specification number, applicable detection range, response change information, and data index information. This information is then integrated and correlated, binding product specification information with the corresponding calibration parameter range. This allows for rapid matching of corresponding data based on the current detection object. Furthermore, binding the detection reference point with the reference response data enables the determination of online detection parameters. The reference state during the process binds multiple calibration nodes within the calibration parameter range to the corresponding response data, enabling reverse thickness lookup based on the actual detection response. For example, data in the 92μm to 108μm range can be integrated as follows: 92μm corresponds to a phase of 134° and a potential of 2.2V; 95μm corresponds to a phase of 140° and a potential of 2.35V; 100μm corresponds to a phase of 150° and a potential of 2.5V; 105μm corresponds to a phase of 160° and a potential of 2.65V; and 108μm corresponds to a phase of 166° and a potential of 2.75V. Simultaneously, 100μm is bound as the detection reference point. After data integration, the above data is saved according to a preset storage format, such as as a detection call data record in the equipment control system. This record includes the product specification number, calibration parameter range, detection reference point, reference response data, local response data, and corresponding version information. When online detection begins, the detection call data is directly read according to the current product identification information, and compared with the reference response data in the real-time acquired coherent response data to quickly determine the thickness deviation. For example, if the online detection obtains a coherent response of 153°, and the 100μm reference phase in the call data is 150°, the corresponding thickness change is determined according to the phase change trend in the range of 92μm to 108μm, and the actual detected thickness is output. This can convert complex full-cycle calibration results into a data structure suitable for rapid on-site retrieval, reducing online calculations and improving detection efficiency. At the same time, it ensures that the detection process is always based on verified calibration data, improving the stability and maintainability of the single-frequency terahertz coherent phase thickness measurement system during long-term operation.

[0036] In a preferred embodiment, during the actual detection process, the detection parameters of the object under test are determined by calling the corresponding detection call data based on the current object being detected and obtaining the offset information between the coherent response data and the detection reference point, including: S701. Obtain the object identification information corresponding to the current detection object, and determine the detection call data that matches the current detection object based on the object identification information; S702. Obtain the detection reference point and corresponding reference response data from the detection call data, and determine the detection reference state corresponding to the current detection object; S703: Control the single-frequency terahertz coherent phase thickness measurement system to detect the current object and obtain the coherent response data corresponding to the current object; S704. Compare the coherent response data with the reference response data in the detection reference state to determine the offset information between the coherent response data and the detection reference point. S705. Based on the offset information and detection call data, determine the detection parameters corresponding to the current detection object.

[0037] As described in steps S701 to S705 above, during the actual testing process, it is necessary to determine the type, specifications, and corresponding testing requirements of the object currently entering the testing area. This allows for the selection of data matching the current object from multiple generated testing call data sets. Object recognition information corresponding to the current object is obtained through an object recognition device located at the testing station. This object recognition device can include an industrial camera, barcode recognition equipment, RFID equipment, or a production system data interface, and is configured according to the production management method of the object being tested. For example, in a lithium electrode sheet thickness testing scenario, an industrial camera located at the testing entrance can capture images of the electrode sheet currently entering the testing area, while simultaneously controlling the production line... The system acquires the product batch number, material type number, and process specification number corresponding to the current electrode. The electrode image is used to identify the appearance category of the object being inspected; the product batch number is used to associate the batch with the corresponding testing standard; the material type number is used to determine the terahertz propagation characteristics of the corresponding material; and the process specification number is used to determine the target inspection thickness range. An industrial camera acquires two-dimensional image data of the object being inspected. Edge extraction, contour localization, and feature region recognition are performed on the image data to obtain the object's dimensional characteristics, such as a length of 1000 mm and a width of 300 mm. Simultaneously, the system reads the corresponding production parameter information of the object being inspected through the production control interface, such as determining that the object belongs to the positive electrode category. For sheet products with a target thickness range of 80μm to 120μm, image recognition results are combined with production parameter information to form object recognition information corresponding to the current detection object. After completing the full-cycle calibration, multiple detection call data are stored. Each detection call data corresponds to a detection range or detection object type. For example, detection call data A corresponds to the detection of positive electrode sheets with a thickness range of 50μm to 100μm, detection call data B corresponds to the detection of negative electrode sheets with a thickness range of 100μm to 200μm, and detection call data C corresponds to the detection of composite materials with a thickness range of 200μm to 500μm. After obtaining the object recognition information of the current detection object, the corresponding material category is determined according to the material type number, and then according to the target... The thickness range determines the corresponding detection range. Finally, data covering the current detection range is selected from existing detection data as the matching result. For example, if the object identification information indicates that the current detection object is a cathode sheet with a target thickness range of 80μm to 120μm, it is determined that detection data A is applicable to a range of 50μm to 100μm, which cannot completely cover the 120μm thickness range. Subsequently, it is determined that detection data B is applicable to a range of 100μm to 200μm, but the material type does not match. Ultimately, based on the detection range coverage and material type consistency, detection data covering 80μm to 150μm and corresponding to the cathode sheet material is selected as the data corresponding to the current detection object. If multiple data meet the conditions...The system further filters data based on its generation time, most recent review time, and current operating status. For example, it prioritizes data with the longest validity period and the most recent successful review as the current data to be called. This allows for automatic matching of the corresponding data based on the actual object being tested, avoiding measurement deviations caused by different objects sharing the same calibration data, improving the accuracy of the calculated testing parameters, reducing manual switching of calibration parameters, and enhancing the automated testing capability of the single-frequency terahertz coherent phase thickness measurement system in continuous production processes. The detection call data includes information such as the detection reference point generated by the full-cycle calibration process, the reference response data corresponding to the detection reference point, and the calibration parameter range associated with the detection reference point. The detection reference point represents the position on the calibration curve corresponding to the target parameter of the current detection object. The reference response data represents the standard coherent response result output by the single-frequency terahertz coherent phase thickness measurement system at that detection reference point position. For example, when detecting electrodes with a thickness range of 80μm to 120μm, the detection call data records the corresponding detection reference point at a thickness of 100μm, with a corresponding coherent response value of 0.65°. This detection reference point is also recorded. The corresponding effective detection range is 90μm to 110μm. The data index information in the detection call data is read, and the corresponding data storage area is located based on the detection call data number corresponding to the current detection object. Then, the detection reference point information, reference response data, and associated calibration parameter range information are read sequentially. The data identifier field of the detection call data is parsed to determine the detection object category corresponding to the current call data. The reference parameter field is read to obtain the standard parameter value corresponding to the detection reference point. For example, if the target thickness parameter is read as 100μm, the response parameter field is then read to obtain the standard coherent response data corresponding to that target thickness parameter. For example, if the coherent response is read... The value is 0.65°, the potential response value is 0.32V, and finally the detection range field is read to obtain the effective detection range corresponding to the detection reference point, for example, 80μm to 120μm. The detection reference state is used to describe the response of the current detection object under standard conditions, including reference detection parameters, reference response data, and reference detection range. The reference detection parameters are derived from the detection reference point in the detection call data; for example, if the target thickness of the current detection object is 100μm, then 100μm is used as the reference detection parameter. The reference response data are derived from the reference response data in the detection call data; for example, 0.65° is used as the system standard response value. The detection range is derived from the calibration parameter range in the detection call data, such as 80μm to 120μm as the current detection range. After the above information extraction is completed, the detection reference point, reference response data, and corresponding detection range are combined to form the detection reference state corresponding to the current detection object. For example, for a detection task of a certain type of positive electrode sheet, if the matching detection call data contains a detection reference point with a thickness of 100μm, a reference response data with a coherent response of 0.65°, and a detection range of 80μm to 120μm, then the corresponding detection reference state is generated, in which the theoretical thickness reference value of the current detection object is recorded as 100μm and the standard coherent response value is 0.With a detection angle of 65° and an allowable detection range of 80μm to 120μm, a detection reference state is established, enabling the real-time coherent response data acquired during actual detection to be compared with a unified standard state. This avoids inconsistencies in the reference benchmark due to different detection objects, different calibration cycles, or different detection ranges, thereby improving the stability and repeatability of the detection parameter calculation process. Based on the detection range and location requirements in the detection reference state, the position of the detection mechanism is adjusted so that the current object to be detected enters the terahertz detection region. For example, for continuously conveyed battery electrodes, the conveying mechanism moves the electrodes between the terahertz transmitting module and the coherent receiving module, aligning the detection region with the area to be tested on the electrodes. The terahertz transmitting module is then activated to output a single-frequency terahertz signal, such as a continuous wave signal with a frequency of 300 GHz. After the terahertz signal irradiates the surface of the current object to be detected, the propagation path changes due to differences in the internal structure and thickness of the material, causing a phase change in the returned signal. The coherent receiving module receives the terahertz signal after passing through the object under test and performs coherent processing on the received signal and the reference signal to obtain the corresponding coherent interference information. The signal acquisition module samples the coherent output signal and converts the analog response signal into digital response data. For example, in one detection process, 100 consecutive coherent sampling results are acquired, with each sampling interval being 1 ms, resulting in 100 sets of coherent response data. The 100 sets of data are averaged to reduce the influence of random noise, obtaining the stable coherent response corresponding to the current object to be detected. The response results are processed by reading the detection position code corresponding to the current detection object, determining the detection area, and controlling the terahertz transmitting module, coherent receiving module, and signal acquisition module to work synchronously. Next, the response information of the terahertz signal after passing through the current detection object is acquired, including phase change, amplitude change, and signal stability parameters. Finally, the acquired data is filtered, for example, removing sampling points with amplitude abnormalities exceeding 20% ​​of the normal range, and the remaining valid sampling data is averaged to obtain the final coherent response data. For example, five sets of valid coherent response data are acquired. The values ​​are 0.63°, 0.64°, 0.65°, 0.66°, and 0.64° respectively. By adding the five sets of data, we get 3.22°. Then, we divide 3.22° by 5 to get 0.644°. We use 0.644° as the coherent response data corresponding to the current detection object, so as to realize the digital acquisition of the actual state of the current detection object. It can obtain real-time response information reflecting the thickness change of the measured object. At the same time, through synchronous acquisition and data processing, we reduce the impact of environmental vibration, signal fluctuation and equipment status changes on the detection results, and improve the stability and accuracy of the detection process. The reference response data is read from the established detection reference state. This reference response data represents the theoretical response value that the current detection object should correspond to under the target detection parameters. For example, when the target thickness corresponding to the detection reference point is 100 μm, the standard coherent response value corresponding to this thickness is recorded as 0.650° in the detection call data. The actual coherent response data obtained from the current detection object is read, for example, the actual coherent response data obtained from the current detection object is 0.674°. The real-time coherent response data is compared with the reference response data to determine the response difference between the two. Based on the response parameters in the detection call data... The comparison object is determined by the type of data. For example, if coherent response is used as the parameter to characterize thickness variation, the reference coherent response value and the currently detected coherent response value are read. Then, they are converted according to a unified data unit to ensure that the two response data are in the same dimension, for example, both are converted to ° units. The change between the two is calculated, that is, by comparing the difference between the coherent response data of the current detected object and the reference response data, the response offset is obtained. For example, if the coherent response data of the current detected object is 0.674° and the reference response data is 0.650°, subtracting 0.650° from 0.674° yields 0.024°. 0.024° represents the current detected... The response offset of the object being measured relative to the detection reference point is determined during the full-cycle calibration process, based on the response changes corresponding to different standard parameters. For example, during the calibration phase, the response values ​​were 0.620° for a thickness of 90 μm, 0.650° for 100 μm, and 0.680° for 110 μm. Therefore, it can be determined that around 100 μm, a response change of 0.030° corresponds to a thickness change of 10 μm. When the current object being measured exhibits a response offset of 0.024°, based on the aforementioned calibration data, the corresponding thickness change is approximately 8 μm. Therefore, the thickness change is determined based on the response offset. The parameters corresponding to the reference point are offset in direction and degree. For example, if the current detection response value is greater than the reference response value, it is determined that the thickness of the current object relative to the detection reference point has increased; if the current detection response value is less than the reference response value, it is determined that the thickness of the current object relative to the detection reference point has decreased. The coherent response data obtained in real time is converted into offset information relative to the standard reference state, so that the detection process no longer depends on the single absolute measurement result, but compares it with the reference state obtained after full cycle calibration. This reduces the impact of equipment zero point changes, environmental disturbances and long-term operation drift on the detection results, and improves the accuracy and stability of detection parameter calculation. The system reads the currently invoked detection data, which includes the standard detection parameters corresponding to the detection reference point, the detection reference response data, and the calibration parameter range near the detection reference point. For example, the current detection data records that the thickness parameter corresponding to the detection reference point is 100μm, the reference response data is 0.650°, and the response change data in the range of 80μm to 120μm near the detection reference point is also recorded. Based on the obtained offset information, the direction and degree of parameter change of the current detection object relative to the detection reference point are determined. For example, by comparison, it is found that the response value of the current detection object is 0.024° higher than the reference response value. Based on the response change trend in the calibration curve, the system determines... This offset indicates an increase in thickness. The calibration data near the detection reference point is read from the detection call data, and the corresponding parameter position is found based on the response offset. For example, if the detection call data record includes a response value of 0.650° for a thickness of 100μm and 0.680° for a thickness of 110μm, then a response increase of 0.030° corresponds to a thickness increase of 10μm. If the current response increase is 0.024°, then the thickness increase is determined to be approximately 8μm through proportional correlation. It is determined that a 10μm thickness change produces a 0.030° response change, and the current response change is determined to be 0.024°. By comparing the ratio of 0.024° to 0.030°, the current... The response change accounts for approximately 80% of the complete range of change. Multiplying 10μm by 80% yields a thickness offset of approximately 8μm. Therefore, adding 8μm to the detection reference point of 100μm results in a current detection object thickness of approximately 108μm. If the response value during detection is lower than the reference response value, the detection parameters are corrected in the opposite direction. For example, if the reference thickness is 100μm, a response offset corresponds to a thickness reduction of 5μm, resulting in a final detection result of 95μm. To improve the reliability of the detection results, detection parameters can be calculated separately for multiple consecutive detection positions, and consistency judgment can be performed on the continuous detection results. For example, continuous detection of 1... If thickness parameters of 108μm, 107μm, 109μm, and 108μm are obtained at 0 locations, the detection result is considered stable. If a significant deviation occurs, such as a detection result of 150μm at a certain point, the detection data is re-called for verification or abnormal results are removed. This achieves the conversion from terahertz coherent response changes to actual thickness parameters, enabling the single-frequency terahertz coherent phase thickness measurement system to perform rapid online detection using full-cycle calibration results, improving the accuracy of thickness measurement results. By calling pre-generated detection data, recalibration is avoided for each detection, shortening the detection time and improving detection efficiency and system adaptability in continuous production environments.

[0038] In a preferred embodiment, when preset verification conditions are met, the detection call data is verified using standard samples, and the processing method corresponding to the detection call data—whether to retain it for use, compensate for adjustments, or re-execute the full-cycle calibration—is determined based on the verification results, including: S801. Obtain the running status information corresponding to the detection call data, and determine whether the preset review conditions are met based on the running status information; When the running status information meets the preset review conditions, the standard sample corresponding to the detection call data is called, the standard sample is tested, and the review response data corresponding to the standard sample is obtained. S802. Obtain the baseline response data from the detection call data, compare the verification response data with the baseline response data, and determine the verification offset information corresponding to the detection call data. S803. Determine the valid status of the detection call data based on the verification offset information, and determine the corresponding processing method for the detection call data based on the valid status.

[0039] As described in steps S801 to S803 above, the running status information corresponding to the detection call data is obtained. This running status information reflects the stability of the current detection call data during actual use, including call count information, usage time information, detection environment information, equipment status information, and historical detection deviation record information. The call count information is obtained by reading the data access record corresponding to the detection call data. For example, each time the system completes an online thickness detection, a count is added to the call record. When the cumulative call count of detection call data A reaches 10,000, the current call count is recorded as 10,000. The usage time information is obtained by reading the generation time of the detection call data and the current system time. For example, the detection call data... If the generation time of A is January 1st of a certain year, and the current detection time is July 1st of a certain year, then the data used for this detection has been continuously used for 6 months. Environmental information is obtained through environmental sensors installed in the detection area, such as obtaining the current temperature of 28℃ and humidity of 55%, and comparing it with the environmental parameters recorded during initial calibration. Equipment status information is obtained by reading the internal operating parameters of the single-frequency terahertz coherent phase thickness measurement system, such as obtaining terahertz transmission power of 95%, received signal strength of 98%, and the phase demodulation module operating normally. Historical detection deviation records are obtained by reading detection deviation data obtained previously using standard samples or manual verification, such as recording the maximum deviation that occurred during the last 20 verifications. The value is 2μm. Based on the acquired operating status information, it is determined whether the preset verification conditions have been met. These preset verification conditions are used to determine whether the current detection data needs to be re-verified. Multiple conditions can be set according to actual application needs. For example, the first verification condition can be that the number of detection calls reaches a preset number, such as a cumulative call count of 5000 times; the second verification condition can be that the continuous running time of the detection data exceeds a preset time, such as a continuous running time exceeding 180 days; the third verification condition can be that the historical detection deviation exceeds the allowable range, such as a deviation exceeding ±2μm for three consecutive detection results; and the fourth verification condition can be that the equipment operating status is abnormal, such as a decrease in terahertz transmission power exceeding 10% compared to the initial calibration state. The status information is checked item by item to determine if the above conditions are met. For example, if the current detection data has been used a cumulative total of 6200 times, exceeding the 5000-times review threshold, then the review condition is met. Alternatively, if the current detection data has been used for 200 days, exceeding 180 days, the review condition is also met. If multiple conditions exist simultaneously, the review process is triggered as long as any one condition is met. When the running status information meets the preset review conditions, the standard sample corresponding to the detection data is called for review testing. The standard sample comes from the standard sample used to establish the detection data during the full-cycle calibration process. It has known standard parameters, such as a standard sample thickness of 100μm, and the standard sample has already obtained corresponding standard response data in the initial calibration stage.Based on the sample association information in the detection call data, the corresponding standard sample number is determined. For example, if detection call data A is associated with the standard sample number THZ-100, the sample transport mechanism is controlled to move the standard sample to the terahertz detection area, ensuring the standard sample is in the same position as the actual detection. Then, the single-frequency terahertz coherent phase thickness measurement system is activated to detect the standard sample, collecting the coherent response information of the terahertz signal after passing through the standard sample. For example, a 300 GHz terahertz signal is transmitted to the standard sample, and the coherent signal after passing through the standard sample is received. 100 sets of coherent response data are continuously collected. For values ​​such as 0.648°, 0.651°, and 0.649°, the collected data undergoes stabilization processing to remove abnormal fluctuations, such as removing outliers exceeding ±5% of the average value. The remaining valid data is then averaged to obtain the verification response data corresponding to the standard sample. For example, if the average of the valid sampled data yields 0.650°, then 0.650° is used as the verification response data corresponding to the current detection data. This allows for proactive verification of the current state of the detection data after long-term use, preventing gradual deviation of detection results due to system drift and improving the long-term stability of online thickness measurement. The baseline response data is read from the detection call data. This baseline response data originates from the response results corresponding to the detection baseline points determined during the full-cycle calibration phase. The corresponding data storage area is located based on the current detection call data number. The detection baseline point field is read to determine the target detection parameter, e.g., 100μm. The response field corresponding to this detection baseline point is then read to obtain the baseline response data, e.g., 0.650°. The data types and units of the two sets of data are then compared. If both are coherent response data, they are directly compared. If there are unit differences, a unified conversion is performed. For example, if the verification response data is 0.657° and the baseline response data is 0.650°, then 0.657° minus 0.650° yields 0.007°. This result indicates that the current detection call data has a 0.007° response offset relative to the initial calibration state. Based on the detection... The calibration curve information in the test call data is used to convert the response offset into the corresponding parameter offset. For example, if the calibration relationship around 100μm is recorded in the test call data as a response value of 0.650° for a thickness of 100μm and 0.665° for a thickness of 105μm, then an increase of 5μm in thickness will result in a change of 0.015°. If the current verification offset is 0.007°, it means that the current response change corresponds to a thickness offset of approximately 2.3μm. This determines the verification offset information corresponding to the test call data, including the response offset amount, parameter offset direction, and parameter offset degree. For example, if the verification offset information is that the response offset increases by 0.007°, the thickness direction increases, and the expected parameter offset is approximately 2.3μm, then the current actual response state of the device can be quantitatively compared with the historical calibration state, thereby accurately determining whether the test call data still meets the online testing requirements. The permissible deviation condition corresponding to the detection call data is read. This permissible deviation condition is derived from the actual thickness measurement accuracy requirements and calibration experiment results. For example, for an application scenario where the thickness detection accuracy requirement is ±2μm, the permissible parameter offset range for the detection call data can be set to -2μm to +2μm. When the verified parameter offset is within this range, the detection call data is considered still valid. When the verified offset information shows that the thickness change corresponding to the current detection call data response offset is 1μm, since 1μm is less than the 2μm permissible range, the validity status of the detection call data is judged to be normal. If the thickness change corresponding to the verified offset is 3μm, it exceeds the permissible range, and the validity status of the detection call data is judged to be abnormal. The corresponding processing method is determined according to the validity status of the detection call data. When the validity status of the detection call data is normal, the current detection call data remains unchanged and continues to be used for online detection. For example, if the verification result shows that the thickness error of the standard sample detection is 1μm, the current detection call data continues to be called. When the detection call data is stored... When there is a slight offset but correction is still possible, such as a thickness change of 2.5 μm corresponding to the verification offset, but the overall response trend of the detection call data remains unchanged, the reference response data in the detection call data is adjusted according to the verification offset. For example, the original reference response data of 0.650° is corrected to 0.657° to ensure that subsequent detections remain accurate. When the offset of the detection call data is large or cannot be recovered by correction, such as a thickness change exceeding 5 μm corresponding to the verification offset, or multiple consecutive verifications exceeding the allowable range, the current detection call data is deemed invalid. The full-cycle calibration process needs to be re-executed, standard sample response data is re-acquired, and new detection call data is generated. This enables dynamic maintenance of the detection call data during long-term operation, timely detection of calibration drift problems based on actual operating conditions, and the adoption of corresponding processing strategies to maintain detection accuracy. This avoids overall deviation in thickness measurement results due to calibration data failure, and improves the long-term reliability and automated maintenance capability of the single-frequency terahertz coherent phase thickness measurement system.

[0040] And, the full-cycle calibration terminal of the single-frequency terahertz coherent phase thickness measurement system, including: One or more processors; A storage device on which one or more programs are stored; When one or more programs are executed by one or more processors, the one or more processors implement a full-cycle calibration method for a single-frequency terahertz coherent phase thickness measurement system.

[0041] The above description is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained in this invention are implemented according to conventional methods in the art unless otherwise specified or limited.

Claims

1. A full-cycle calibration method for a single-frequency terahertz coherent phase thickness measurement system, characterized in that, include: Acquire the calibration status information of the single-frequency terahertz coherent phase thickness measurement system, perform benchmark calibration on the single-frequency terahertz coherent phase thickness measurement system, and establish the initial measurement benchmark corresponding to the calibration process. Configure standard samples for calibrating the single-frequency terahertz coherent phase thickness measurement system, and determine the calibration parameter sequence based on the standard parameters corresponding to the standard samples; The standard sample is periodically scanned within the effective calibration interval according to the calibration parameter sequence to obtain coherent response data and standard parameter data corresponding to different calibration positions, forming periodic calibration acquisition data; The periodic calibration data is processed for data quality, and valid calibration data that meets the preset quality conditions is selected. A calibration curve is generated based on the effective calibration data. According to the standard parameters of the target detection object, a detection reference point is determined on the calibration curve, and the reference response data corresponding to the detection reference point is obtained. Based on the calibration curve, the detection reference point, and the corresponding reference response data, generate the detection call data; In the actual testing process, the corresponding testing call data is called according to the current testing object, and the offset information between the coherent response data and the testing reference point is obtained to determine the testing parameters of the tested object; When the preset verification conditions are met, the detection call data is verified using standard samples, and the processing method corresponding to the detection call data is determined based on the verification results: retain for use, compensate for adjustment, or re-execute full-cycle calibration.

2. The full-cycle calibration method for the single-frequency terahertz coherent phase thickness measurement system according to claim 1, characterized in that, Acquire the calibration status information of the single-frequency terahertz coherent phase thickness measurement system, perform benchmark calibration on the single-frequency terahertz coherent phase thickness measurement system, and establish the initial measurement benchmark corresponding to the calibration process, including: Acquire the calibration status information corresponding to the single-frequency terahertz coherent phase thickness measurement system, wherein the calibration status information includes terahertz transmission status information, reception status information, signal acquisition status information and phase demodulation status information; Determine whether the single-frequency terahertz coherent phase thickness measurement system meets the preset calibration conditions based on the equipment status information. When the single-frequency terahertz coherent phase thickness measurement system meets the preset calibration conditions, the terahertz transmitting module, coherent receiving module, signal acquisition module and phase demodulation module are initially calibrated. Obtain the output response data after the initial calibration is completed, and determine the initial measurement benchmark corresponding to the calibration process based on the output response data.

3. The full-cycle calibration method for the single-frequency terahertz coherent phase thickness measurement system according to claim 1, characterized in that, Configure standard samples for calibrating a single-frequency terahertz coherent phase thickness measurement system, and determine the calibration parameter sequence based on the standard parameters corresponding to the standard samples, including: Obtain the detection requirements corresponding to the object to be tested, and configure standard samples for calibrating the single-frequency terahertz coherent phase thickness measurement system; Obtain the standard parameter information corresponding to the standard sample, and determine the parameter coverage range corresponding to the standard sample based on the standard parameter information; The calibration acquisition range is determined based on the parameter coverage range, and the calibration acquisition range is divided according to the preset scanning rules to obtain the calibration parameters corresponding to multiple calibration positions. The execution order corresponding to each calibration parameter is determined based on multiple calibration parameters and preset scanning rules, and a calibration parameter sequence is generated.

4. The full-cycle calibration method for the single-frequency terahertz coherent phase thickness measurement system according to claim 1, characterized in that, The standard sample is periodically scanned within the effective calibration interval according to the calibration parameter sequence to acquire coherent response data and standard parameter data corresponding to different calibration positions, forming periodic calibration acquisition data, including: Determine the scanning path and scanning nodes corresponding to the standard sample based on the calibration parameter sequence; The standard sample is controlled to move sequentially to each scanning node according to the scanning path within the effective calibration range to obtain the corresponding calibration position; When the standard sample is at each calibration position, acquire the coherent response data output by the single-frequency terahertz coherent phase thickness measurement system. The coherent response data includes phase response data and potential response data, and read the standard parameter data corresponding to the calibration position. Based on the coherent response data and standard parameter data corresponding to each calibration location, periodic calibration acquisition data is generated.

5. The full-cycle calibration method for the single-frequency terahertz coherent phase thickness measurement system according to claim 1, characterized in that, The periodic calibration data is processed for data quality, and valid calibration data that meets preset quality conditions is selected, including: Based on the periodic calibration data, obtain the coherent response data, standard parameter data and corresponding acquisition status information for each calibration position. Based on the acquisition status information, perform data correction processing on the coherent response data and standard parameter data for each calibration position to obtain the corrected calibration data. According to the preset quality evaluation rules, the quality of the calibrated data is evaluated to obtain the data quality evaluation results for each calibration location. Obtain quality matching information between the data quality evaluation results and the preset quality conditions, and determine the calibration data that meets the preset quality conditions as valid calibration data.

6. The full-cycle calibration method for the single-frequency terahertz coherent phase thickness measurement system according to claim 1, characterized in that, A calibration curve is generated based on valid calibration data. According to the standard parameters of the target object, a detection reference point is determined on the calibration curve, and the reference response data corresponding to the detection reference point is obtained, including: Based on the standard parameter data and coherent response data in the effective calibration data, a corresponding mapping between calibration parameters and response parameters is established, and calibration curves are generated according to the corresponding mapping. Obtain the standard parameter information corresponding to the target detection object, and perform parameter positioning on the calibration curve based on the standard parameter information to determine the calibration position corresponding to the target detection object; The detection reference point is determined based on the calibration position corresponding to the target detection object, and the response parameters corresponding to the detection reference point are obtained. The response parameters are used as the reference response data.

7. The full-cycle calibration method for the single-frequency terahertz coherent phase thickness measurement system according to claim 1, characterized in that, Based on the calibration curve, detection reference points, and corresponding reference response data, detection call data is generated, including: Obtain the position of the detection reference point in the calibration curve, and determine the calibration parameter range corresponding to the detection reference point from the calibration curve; Based on the calibration parameter range, obtain the benchmark response data corresponding to the detection benchmark point, and determine the benchmark call information; Data is integrated based on calibration parameter ranges, detection benchmarks, and benchmark call information to generate detection call data for use in the online detection phase.

8. The full-cycle calibration method for the single-frequency terahertz coherent phase thickness measurement system according to claim 1, characterized in that, In the actual testing process, the corresponding testing call data is invoked based on the current testing object, and the offset information between the coherent response data and the testing reference point is obtained to determine the testing parameters of the tested object, including: Obtain the object recognition information corresponding to the current detection object, and determine the detection call data that matches the current detection object based on the object recognition information; Obtain the detection reference point and corresponding reference response data from the detection call data, and determine the detection reference state corresponding to the current detection object; The single-frequency terahertz coherent phase thickness measurement system is controlled to detect the current object and acquire the coherent response data corresponding to the current object. The coherent response data is compared with the reference response data in the detection reference state to determine the offset information between the coherent response data and the detection reference point. Based on the offset information and detection call data, determine the detection parameters corresponding to the current detection object.

9. The full-cycle calibration method for the single-frequency terahertz coherent phase thickness measurement system according to claim 1, characterized in that, When the preset verification conditions are met, the detection call data is verified using standard samples, and the processing method corresponding to the detection call data—whether to retain it for use, compensate for and adjust it, or re-execute the full-cycle calibration—is determined based on the verification results, including: Obtain the running status information corresponding to the detection call data, and determine whether the preset review conditions have been met based on the running status information; When the running status information meets the preset review conditions, the standard sample corresponding to the detection call data is called, the standard sample is tested, and the review response data corresponding to the standard sample is obtained. Obtain the baseline response data from the detection call data, compare the review response data with the baseline response data, and determine the review offset information corresponding to the detection call data; The validity status of the detection call data is determined based on the verification offset information, and the corresponding processing method for the detection call data is determined based on the validity status.

10. A full-cycle calibration terminal for a single-frequency terahertz coherent phase thickness measurement system, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When one or more programs are executed by one or more processors, the one or more processors implement the full-cycle calibration method of the single-frequency terahertz coherent phase thickness measurement system according to any one of claims 1 to 9.

Images