Intelligent monitoring system for black titanium surface coloring

Abstract

The invention relates to the technical field of metal surface treatment monitoring, and discloses a black titanium surface coloring intelligent monitoring system, which comprises a spectrum acquisition module, an environment disturbance acquisition module, an environment disturbance comparison module and the like. The spectrum acquisition module dynamically adjusts a sampling frequency and a wavelength range in combination with environmental disturbance parameters; the environment disturbance acquisition module acquires information such as a temperature gradient change rate and an environment humidity fluctuation rate, and the environment disturbance comparison module generates a comprehensive index according to the information and judges an environment disturbance level. And the equipment state acquisition and comparison module monitors internal operation parameters of the system, generates an equipment health index and deals with abnormity. The spectrum and model synchronization module matches and evaluates the coloring quality by using wavelet transform and a pre-training model, and the coloring abnormity analysis module identifies defects. In addition, the system is also provided with a dynamic calibration module. The system can accurately monitor black titanium surface coloring, and the production efficiency and the product quality are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of metal surface treatment monitoring, and specifically to an intelligent monitoring system for black titanium surface coloring. Background Technique

[0002] In the field of metal surface treatment, black titanium is widely used in many industries such as architectural decoration, kitchenware manufacturing, and electronic product casings due to its unique appearance and excellent performance. With the continuous increase in market demand for black titanium products and the increasingly stringent requirements for product quality, the control of black titanium surface coloring quality has become a key link. However, there are many problems in the current monitoring of black titanium surface coloring process, which seriously restricts the improvement of product quality and the development of the industry.

[0003] From the perspective of environmental factors, the existing monitoring means are difficult to effectively cope with the influence of complex and changeable environments on black titanium surface coloring. Temperature and humidity are important environmental factors affecting the spectral stability of black titanium surface. In the actual production environment, the fluctuation of temperature will cause thermal expansion of black titanium materials, which will in turn affect their surface microstructure and change the light reflection characteristics. For example, in some large metal processing workshops, the temperature difference between day and night is relatively large. If the monitoring strategy is not adjusted in time, the collected spectral data may deviate greatly and cannot accurately reflect the true coloring situation of the black titanium surface. In terms of humidity, the change of environmental humidity will form water films with different thicknesses on the black titanium surface, accelerating or inhibiting the surface oxidation process, which will also interfere with the spectral stability. Traditional monitoring systems often ignore the dynamic changes of these environmental factors, or can only record simple environmental parameters, and cannot adjust the monitoring process in real time based on these parameters, resulting in a significant reduction in the reliability of monitoring results.

[0004] In terms of the operation status of the equipment itself, there are also obvious deficiencies in the existing monitoring systems. As the core equipment for obtaining black titanium surface spectral data, spectrometers often have problems such as calibration offset, data sampling period jitter, and signal baseline drift. For example, a spectrometer that has been used for a long time may have a wavelength baseline shift, making the collected wavelength data inaccurate, which will in turn affect the judgment of the black titanium surface coloring quality. The instability of the data sampling period will cause the collected data to not accurately reflect the real-time state of the black titanium surface, resulting in information loss or redundancy. The signal baseline drift will introduce additional noise and interfere with the analysis of spectral data. At present, most monitoring systems lack real-time monitoring and effective analysis of these equipment operation parameters, and cannot detect potential equipment failure hazards in time. When serious problems occur in the equipment, repairs or replacements are carried out, which not only increases production costs but also causes production interruptions and affects production efficiency.

[0005] In terms of coloring quality assessment, traditional methods mainly rely on manual experience or simple visual inspection. This method is highly subjective, has low precision and low efficiency. Manual judgment is greatly affected by factors such as the professional level, working state and environment of the inspectors, and it is difficult to ensure the consistency and accuracy of the inspection results. For some minor coloring defects or film thickness deviations, manual inspection is very likely to miss them. Moreover, with the expansion of production scale, the efficiency of manual inspection far cannot meet the production requirements. In addition, some existing instrument-based detection methods can often only detect a single indicator, such as only detecting whether the color meets the standard, and cannot comprehensively evaluate multiple key indicators such as coloring uniformity and film thickness deviation, and cannot provide a comprehensive and effective quality control basis for the production process. Summary of the Invention

[0006] The purpose of the present invention is to provide an intelligent monitoring system for black titanium surface coloring based on a certain method to solve the problems raised in the above background technology.

[0007] To achieve the above purpose, the present invention provides the following technical solutions: An intelligent monitoring system for black titanium surface coloring based on a certain method, including a spectral acquisition module, and also including an environmental disturbance acquisition module, an environmental disturbance comparison module, a device state acquisition module, a device state comparison module, a spectrum and model synchronization module, and a coloring anomaly analysis module;

[0008] The spectral acquisition module is used to perform real-time spectral reflectivity detection on the black titanium surface, and dynamically adjust the spectral sampling frequency and wavelength range in combination with environmental disturbance parameters;

[0009] The environmental disturbance acquisition module is used to obtain external environmental parameters that affect spectral stability, including temperature gradient change rate information and environmental humidity volatility information;

[0010] The environmental disturbance comparison module is used to compare the collected temperature gradient change rate and humidity volatility with a preset threshold one to generate an environmental disturbance comprehensive index; if the environmental disturbance comprehensive index exceeds threshold one, the system determines that the environmental disturbance level is high risk, suspends spectral acquisition and starts an anti-interference protocol; if it is lower than or within the range of threshold one, it is determined to be low risk, and the device state acquisition is started;

[0011] The device state acquisition module is used to monitor the operating parameters of internal sensors and processing units of the system, including spectrometer calibration offset rate, data sampling period jitter rate, signal baseline drift rate;

[0012] The device state comparison module generates a device health index by analyzing the calibration offset rate, sampling period jitter rate and baseline drift rate, and compares it with a preset threshold two; if the device health index exceeds threshold two, it is determined that the device state is abnormal and parameter reset is triggered; if it is lower than or within the range of threshold two, spectrum and model synchronization is started;

[0013] The spectral synchronization module with the model decomposes the spectral reflectance data by wavelet transform and performs time-frequency domain matching with the pre-trained coloring quality evaluation model;

[0014] The coloring anomaly analysis module is used to identify the deviation between the spectral features and the model benchmark during the synchronization process. If device status anomalies or high-risk environmental disturbances are detected, the current processing flow is frozen and the system parameters are re-initialized.

[0015] Preferably, the temperature gradient change rate information includes the thermal expansion covariance index; the environmental humidity volatility information includes the water film thickness change index;

[0016] The calculation formula for the thermal expansion covariance index is:

[0017]

[0018] where C T represents the thermal expansion covariance index, reflecting the coupling effect of temperature fluctuations on the deformation of the black titanium surface; ΔT i is the temperature gradient change amount at the i-th sampling; μ T and σ T are the mean and standard deviation of the temperature change respectively; α is the thermal expansion coefficient adjustment factor;

[0019] The calculation formula for the water film thickness change index is:

[0020]

[0021] where W h represents the water film thickness change index, reflecting the influence of humidity fluctuations on surface oxidation; H i is the humidity fluctuation amplitude at the i-th sampling; is the mean of the humidity fluctuations; β is the humidity sensitivity coefficient.

[0022] Preferably, the environmental disturbance comprehensive index is generated by the non-linear superposition of the thermal expansion covariance index and the water film thickness change index, and the calculation formula is:

[0023] E d = γ·(C T ·W h + ∥C T - W h ∥)

[0024] where E d is the environmental disturbance comprehensive index; γ is the environmental coupling weight coefficient; ∥·∥ represents the Manhattan distance between the two indexes.

[0025] Preferably, the spectrometer calibration offset rate includes a wavelength baseline offset index; the data sampling period jitter rate includes a timing phase jitter index; the signal baseline drift rate includes a noise covariance index;

[0026] The calculation formula for the wavelength baseline offset index is:

[0027]

[0028] where B λ is the wavelength baseline offset index; λ i is the wavelength deviation at the i-th sampling point; λ0 is the calibrated reference wavelength; δ is the wavelength calibration sensitivity coefficient;

[0029] The calculation formula for the timing phase jitter index is:

[0030]

[0031] where J t is the timing phase jitter index; t i is the time interval of the i-th sampling; is the average sampling period; ∈ is the time jitter correction factor;

[0032] The calculation formula for the noise covariance index is:

[0033]

[0034] where N s is the noise covariance index; S i is the signal baseline drift at the i-th sampling; is the mean baseline drift; ζ is the noise suppression coefficient.

[0035] A smart monitoring system for black titanium surface coloring according to claim 4, characterized in that:

[0036] The device health index is generated by fusing the entropy values of the wavelength baseline offset index, the timing phase jitter index, and the noise covariance index, and the calculation formula is:

[0037]

[0038] where D h is the device health index; η is the device status fusion coefficient.

[0039] Preferably, the wavelet transform decomposition uses the Morlet wavelet basis function to extract multi-scale frequency domain features from the spectral reflectance data, generating low-frequency approximation components and high-frequency detail components; inputting each component into the coloring quality evaluation model to calculate the similarity distance from the standard coloring spectrum.

[0040] Preferably, the coloring quality evaluation model is constructed based on the principal component analysis algorithm, and the Mahalanobis distance matching is performed between the spectral feature vectors after dimensionality reduction and the pre-stored standard sample library to determine the coloring uniformity and the film thickness deviation.

[0041] Preferably, it further includes a dynamic calibration module for adaptively adjusting the integration time and gain parameters of the spectrometer according to the environmental disturbance comprehensive index and the equipment health index.

[0042] Preferably, the coloring anomaly analysis module uses a support vector machine classifier to perform hyperplane division on the spectral feature vectors, thermal imaging data, and roughness data, identify the types of coloring defects, and output anomaly codes.

[0043] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0044] The system uses the environmental disturbance acquisition module to obtain external environmental parameters affecting spectral stability in real time, such as the temperature gradient change rate information (thermal expansion covariance index) and the environmental humidity volatility information (water film thickness change index). Using these parameters, the spectral acquisition module can dynamically adjust the spectral sampling frequency and wavelength range. When the environmental temperature gradient change rate is large or the humidity volatility is high, the system will automatically increase the sampling frequency to ensure more timely capture of spectral changes; at the same time, reasonably adjust the wavelength range to avoid interference of environmental factors on spectral data, so as to ensure that the collected spectral data truly and accurately reflects the coloring situation of the black titanium surface. The environmental disturbance comparison module calculates the environmental disturbance comprehensive index and compares it with a preset threshold. When the environmental disturbance level is high-risk, it suspends spectral acquisition and starts an anti-interference protocol, further ensuring the reliability of monitoring data and effectively solving the problem that traditional monitoring systems are greatly affected by environmental factors.

[0045] The device status acquisition module monitors the operating parameters of internal sensors and processing units of the system in real time, including the spectrometer calibration offset rate (wavelength baseline offset index), data sampling period jitter rate (timing phase jitter index), signal baseline drift rate (noise covariance index), etc. The device status comparison module generates an equipment health index by analyzing these parameters and compares it with a preset threshold II. Once the equipment health index exceeds the threshold, the system determines that the device status is abnormal and triggers parameter reset, enabling the device to quickly return to the normal operating state. This mechanism can timely detect potential equipment failure hazards, avoid serious impacts of equipment failures on monitoring results, reduce production interruptions and product quality degradation caused by equipment problems, while reducing equipment maintenance costs and improving production efficiency.

[0046] The spectrum and model synchronization module uses wavelet transform to decompose the spectral reflectance data, and matches it in the time and frequency domain with the pre-trained coloring quality assessment model based on the principal component analysis algorithm. The model can accurately determine the coloring uniformity and film thickness deviation by matching the spectral feature vector after dimension reduction with the pre-stored standard sample library by Mahalanobis distance. Compared with the traditional method that relies on manual experience or simple visual inspection, the evaluation method of this system is more objective, accurate and efficient. It not only greatly improves the accuracy and consistency of detection, but also can quickly process large amounts of data to meet the detection needs of large-scale production. At the same time, it can detect tiny coloring defects, provide strong support for the optimization of the production process, and help improve the overall quality of the product.

[0047] The dynamic calibration module adaptively adjusts the integration time and gain parameters of the spectrometer according to the comprehensive index of environmental disturbance and the equipment health index, further optimizing the spectrum acquisition process to ensure that high-quality spectrum data can be obtained in different environments and equipment conditions. The coloring anomaly analysis module uses a support vector machine classifier to perform hyperplane division on the spectral feature vector, thermal imaging data, and roughness data, and can accurately identify the type of coloring defects and output anomaly codes. This enables operators to quickly understand the specific abnormal conditions of the black titanium surface, take targeted measures to adjust and improve in a timely manner, effectively reduce the scrap rate, and improve the economic benefits of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a working principle diagram of the monitoring system of the present invention;

[0049] Figure 2 Flowchart for the generation and determination of the comprehensive index of environmental disturbance;

[0050] Figure 3 Flowchart for generating and determining equipment health index.

[0051] Figure 4 Flow chart of the principle of synchronous processing of spectrum and model.

[0052] Figure 5 Flowchart showing the coloring exception parsing principle. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] See also Figures 1-5, the present invention provides a technical solution: based on an intelligent monitoring system for black titanium surface coloring, the system includes:

[0055] Spectrum acquisition module: It performs real-time spectral reflectivity detection on the black titanium surface. During the detection process, it dynamically adjusts the spectral sampling frequency and wavelength range in combination with environmental disturbance parameters. For example, when the environmental disturbance parameters indicate that the environmental change is relatively drastic, the sampling frequency is appropriately increased to more timely capture the changes in the spectrum of the black titanium surface; at the same time, according to the environmental factors that may affect the spectrum, the wavelength range is reasonably adjusted to ensure that the collected spectral data can accurately reflect the coloring situation of the black titanium surface.

[0056] Environmental disturbance acquisition module: It is responsible for obtaining external environmental parameters that affect spectral stability, specifically including temperature gradient change rate information and environmental humidity volatility information. These information will be collected in real time and transmitted to the subsequent modules for processing.

[0057] Environmental disturbance comparison module: It compares the collected temperature gradient change rate and humidity volatility with a preset threshold one. Through a specific calculation method, an environmental disturbance comprehensive index is generated. If the environmental disturbance comprehensive index exceeds threshold one, the system determines that the environmental disturbance level is high risk. At this time, the spectrum acquisition is suspended and the anti-interference protocol is started to avoid the interference of environmental factors on the spectral acquisition data and ensure the accuracy of the data; if it is lower than or within the range of threshold one, it is determined to be low risk, and the device status acquisition module is started.

[0058] Device status acquisition module: It monitors the operating parameters of the internal sensors and processing units of the system, including spectrometer calibration offset rate, data sampling period jitter rate, signal baseline drift rate, etc. These parameters reflect the operating conditions of the internal devices of the system and provide a basis for subsequent judgment of whether the devices are working properly.

[0059] Device status comparison module: It generates a device health index by analyzing the calibration offset rate, sampling period jitter rate and baseline drift rate, and compares it with a preset threshold two. If the device health index exceeds threshold two, it is determined that the device status is abnormal, and parameter reset is triggered to make the device return to the normal working state; if it is lower than or within the range of threshold two, the spectrum and model synchronization module is started.

[0060] Spectrum and model synchronization module: It decomposes the spectral reflectivity data by wavelet transform and performs time-frequency domain matching with a pre-trained coloring quality evaluation model. In this way, the collected spectral data can be compared and analyzed with the model, so as to evaluate the coloring quality of the black titanium surface.

[0061] Color anomaly analysis module: Identify the deviation between the spectral features and the model benchmark during the synchronization process. When detecting abnormal device status or high-risk environmental disturbances, freeze the current processing flow and re-initialize the system parameters to ensure that the system can continuously and stably monitor the coloring of the black titanium surface.

[0062] The present invention will be further described below in conjunction with Embodiments 1 to 5:

[0063] Embodiment 1:

[0064] The environmental disturbance acquisition module obtains external environmental parameters that affect spectral stability, where the temperature gradient change rate information includes the thermal expansion covariance index. The thermal expansion covariance index reflects the coupling effect of temperature fluctuations on the deformation of the black titanium surface, and its calculation formula is:

[0065]

[0066] In the formula, C T represents the thermal expansion covariance index; ΔT i is the temperature gradient change amount of the i-th sampling. In practical applications, temperature data can be collected by a temperature sensor every certain time (such as t0 seconds), and the difference between two adjacent temperature data is the temperature gradient change amount; μ T and σ T are the mean and standard deviation of temperature changes respectively, obtained by statistical calculation of multiple temperature gradient change amount data collected over a period of time; α is the thermal expansion coefficient adjustment factor, and its value is set according to the characteristics of the black titanium material and the actual environmental conditions. For example, for a certain specific black titanium material, in a general industrial environment, after multiple tests, the value of α is determined to be 0.5.

[0067] The environmental humidity volatility information includes the water film thickness change index, and its calculation formula is:

[0068]

[0069] Among them, W h represents the water film thickness change index, reflecting the influence of humidity fluctuations on surface oxidation; H i is the humidity fluctuation amplitude of the i-th sampling, which can be obtained by the humidity sensor collecting humidity data in real time and calculating the fluctuation amplitude between two adjacent data; is the mean value of humidity fluctuations, obtained by averaging the humidity fluctuation amplitude data over a period of time; β is the humidity sensitivity coefficient, determined according to the oxidation characteristics of the black titanium surface in different humidity environments. For example, in an environment with a large change in relative humidity, after experimental determination, the value of β is 0.3.

[0070] The environmental disturbance comprehensive index is generated by the non-linear superposition of the thermal expansion covariance index and the water film thickness change index, and the calculation formula is:

[0071] E d = γ·(C T ·W h + ∥C T - W h ∥)

[0072] Among them, E d is the comprehensive environmental disturbance index; γ is the environmental coupling weight coefficient, and its value comprehensively considers the relative importance of the influence of temperature and humidity on spectral stability. For example, in a certain application scenario, through a large number of experimental verifications, it is determined that the value of γ is 0.6; ∥·∥ represents the Manhattan distance between two indices, that is, the sum of the absolute values of the differences between C T and W h . The comprehensive environmental disturbance index calculated by this formula can more comprehensively reflect the influence degree of environmental factors on spectral acquisition.

[0073] Example 2:

[0074] The device status acquisition module monitors the operating parameters of the internal sensors and processing units of the system. The spectrometer calibration offset rate includes the wavelength baseline offset index, and its calculation formula is:

[0075]

[0076] Among them, B λ is the wavelength baseline offset index; λ i is the wavelength deviation at the i-th sampling point. During the operation of the spectrometer, the wavelength deviation of each sampling point is obtained by comparing with the standard wavelength source; λ0 is the calibrated reference wavelength, which is the standard wavelength value calibrated when the spectrometer leaves the factory; δ is the wavelength calibration sensitivity coefficient, which is set according to the model and accuracy requirements of the spectrometer. For example, for a certain model of spectrometer, its value of δ is 0.2.

[0077] The data sampling period jitter rate includes the timing phase jitter index, and its calculation formula is:

[0078]

[0079] Among them, J t is the timing phase jitter index; t i is the time interval of the i-th sampling, which is obtained by recording the time stamps of each sampling and calculating the difference between the adjacent two sampling times; is the average sampling period, which is obtained by averaging the time intervals of multiple samplings within a period of time; is the time jitter correction factor, which is determined according to the requirements of the system for the sampling time accuracy. For example, in the case of high requirements for the sampling time accuracy, the value is 0.8.

[0080] The signal baseline drift rate includes the noise covariance index, and its calculation formula is:

[0081]

[0082] Among them, N s is the noise covariance index; S i is the signal baseline drift amount of the i-th sampling, which can be obtained by processing the collected signal data to remove the effective signal part and obtain the baseline drift amount; is the average baseline drift, which is obtained by averaging the signal baseline drift amounts over a period of time; ζ is the noise suppression coefficient, which is set according to the anti-noise performance requirements of the system. For example, in a complex noise environment, ζ is taken as 0.7.

[0083] The equipment health index is generated by fusing the entropy values of the wavelength baseline offset index, the timing phase jitter index, and the noise covariance index. The calculation formula is:

[0084]

[0085] Among them, D h is the equipment health index; η is the equipment status fusion coefficient, and its value comprehensively considers the weights of the impacts of various parameters on the equipment health status. For example, after a large number of experiments and data analyses, it is determined that η is taken as 0.4. The equipment health index calculated by this formula can accurately reflect the running health status of the internal equipment of the system.

[0086] Example 3:

[0087] The spectrum and model synchronization module decomposes the spectral reflectance data by wavelet transform. Here, the Morlet wavelet basis function is used to extract the multi-scale frequency domain features of the spectral reflectance data. The Morlet wavelet basis function has good time-frequency localization characteristics and can effectively decompose the different frequency components of the spectral data.

[0088] When performing wavelet transform, the spectral reflectance data is decomposed according to different scales to generate low-frequency approximation components and high-frequency detail components. The low-frequency approximation components reflect the main trends and characteristics of the spectral data, while the high-frequency detail components contain the subtle changes and noise information of the spectral data.

[0089] Input each component into the coloring quality evaluation model to calculate its similarity distance from the standard coloring spectrum. For example, for the spectral reflectance data of a certain black titanium surface, after wavelet transform, the low-frequency approximation component A and the high-frequency detail component D are obtained. Input A and D into the coloring quality evaluation model respectively. The model calculates their differences from the standard coloring spectrum in each characteristic dimension through a specific algorithm, and then obtains the similarity distance. The smaller the similarity distance, the closer the current coloring situation of the black titanium surface is to the standard coloring spectrum, and the higher the coloring quality; on the contrary, the larger the similarity distance, the lower the coloring quality.

[0090] Example 4:

[0091] The coloring quality evaluation model is constructed based on the principal component analysis algorithm. The principal component analysis algorithm is a commonly used data dimensionality reduction method that can reduce the dimensionality of data while retaining the main features of the data, improving the calculation efficiency.

[0092] First, a large amount of spectral data of different black titanium surfaces is collected as a sample set. These samples cover different coloring quality situations, including samples with uniform coloring and those with deviations. The spectral data in the sample set is preprocessed, such as removing outliers and normalizing, to ensure the accuracy and consistency of the data.

[0093] Then, the principal component analysis algorithm is used to perform dimensionality reduction on the preprocessed spectral data to obtain the reduced-dimensional spectral feature vectors. By calculating the covariance matrix of the spectral data, solving for the eigenvalues and eigenvectors, and selecting the main eigenvectors to form a new low-dimensional space, the spectral data is projected into this low-dimensional space to obtain the reduced-dimensional spectral feature vectors.

[0094] The reduced-dimensional spectral feature vectors are matched with the pre-stored standard sample library. The standard sample library stores the spectral feature vectors under different standard coloring conditions and their corresponding coloring uniformity and film thickness information. The Mahalanobis distance can take into account the covariance structure of the data and more accurately measure the similarity between two vectors.

[0095] By calculating the Mahalanobis distance between the reduced-dimensional spectral feature vectors and each vector in the standard sample library, the standard sample with the closest distance is found, and based on the information of this standard sample, the coloring uniformity and film thickness deviation of the current black titanium surface are determined. For example, if the standard sample corresponding to the smallest calculated Mahalanobis distance has good coloring uniformity and the film thickness is within the standard range, it is determined that the coloring quality of the current black titanium surface is good; if the closest standard sample has uneven coloring or a large film thickness deviation, it is determined that there is a problem with the coloring of the current black titanium surface, and the specific deviation type and degree are determined according to the deviation situation of the standard sample.

[0096] Example 5:

[0097] The system also includes a dynamic calibration module, which is used to adaptively adjust the integration time and gain parameters of the spectrometer according to the comprehensive environmental disturbance index and the equipment health index. When the comprehensive environmental disturbance index is relatively high or the equipment health index indicates certain problems with the equipment status, the dynamic calibration module will correspondingly adjust the integration time and gain parameters of the spectrometer. For example, if the comprehensive environmental disturbance index exceeds a certain range, in order to ensure the accuracy of the collected spectral data, the dynamic calibration module appropriately increases the integration time of the spectrometer to improve the signal intensity; at the same time, according to the equipment noise situation reflected by the equipment health index, the gain parameter is adjusted to suppress the influence of noise on the spectral data.

[0098] The coloring anomaly analysis module uses a support vector machine classifier to perform hyperplane partitioning on spectral feature vectors, thermal imaging data, and roughness data. First, a large number of black titanium surface sample data containing different coloring defect types are collected, and these sample data include corresponding spectral feature vectors, thermal imaging data, and roughness data. These data are preprocessed, such as normalization, feature selection, etc., to improve the performance of the classifier.

[0099] Then, the support vector machine classifier is used to train the preprocessed data to find an optimal hyperplane to separate the data of different coloring defect types. In practical applications, when the system collects the spectral feature vectors, thermal imaging data, and roughness data of the black titanium surface, these data are input into the trained support vector machine classifier. The classifier identifies the coloring defect type according to the hyperplane partitioning rule and outputs an anomaly code. For example, the anomaly code "001" indicates uneven coloring, and "010" indicates excessive film thickness deviation, etc. In this way, the operator can quickly understand the coloring anomalies on the black titanium surface and take corresponding measures for adjustment and improvement in a timely manner.

[0100] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise", or any other variation thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device including a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or device.

[0101] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent monitoring system for black titanium surface coloring, including a spectral acquisition module, characterized in that: It also includes an environmental disturbance acquisition module, an environmental disturbance comparison module, a device status acquisition module, a device status comparison module, a spectrum and model synchronization module, and a coloring anomaly analysis module; The spectrum acquisition module is used to perform real-time spectral reflectance detection on the black titanium surface, and dynamically adjust the spectral sampling frequency and wavelength range in combination with environmental disturbance parameters; The environmental disturbance acquisition module is used to obtain external environmental parameters affecting spectral stability, including temperature gradient change rate information and environmental humidity volatility information; The environmental disturbance comparison module is used to compare the collected temperature gradient change rate and humidity volatility with a preset threshold one to generate an environmental disturbance comprehensive index; if the environmental disturbance comprehensive index exceeds threshold one, the system determines that the environmental disturbance level is high-risk, suspends spectrum acquisition and starts an anti-interference protocol; if it is lower than or within the range of threshold one, it is determined to be low-risk, and the device status acquisition is started; The device status acquisition module is used to monitor the operating parameters of internal sensors and processing units of the system, including spectrometer calibration offset rate, data sampling period jitter rate, and signal baseline drift rate; The device status comparison module generates a device health index by analyzing the calibration offset rate, sampling period jitter rate, and baseline drift rate, and compares it with a preset threshold two; If the device health index exceeds threshold two, it is determined that the device status is abnormal and parameter reset is triggered; if it is lower than or within the range of threshold two, spectrum and model synchronization is started; The spectrum and model synchronization module uses wavelet transform to decompose spectral reflectance data and perform time-frequency domain matching with a pre-trained coloring quality evaluation model; The coloring anomaly analysis module is used to identify the deviation between spectral features and model benchmarks during the synchronization process. If device status anomalies or high-risk environmental disturbances are detected, the current processing flow is frozen and system parameters are re-initialized.

2. The intelligent monitoring system for black titanium surface coloring according to claim 1, wherein: The temperature gradient change rate information includes a thermal expansion covariance index; the environmental humidity volatility information includes a water film thickness change index; The calculation formula for the thermal expansion covariance index is: Among them, C T represents the thermal expansion covariance index, reflecting the coupling effect of temperature fluctuations on the deformation of the black titanium surface; ΔT i is the temperature gradient change amount of the i-th sampling; μT and σT are the mean and standard deviation of the temperature change respectively; α is the thermal expansion coefficient adjustment factor; The calculation formula for the water film thickness change index is: Among them, W h represents the water film thickness change index, reflecting the influence of humidity fluctuation on surface oxidation; H i is the humidity fluctuation amplitude of the i-th sampling; is the average humidity fluctuation; β is the humidity sensitivity coefficient.

3. The intelligent monitoring system for black titanium surface coloring according to claim 2, wherein: The environmental disturbance comprehensive index is generated by non-linear superposition of the thermal expansion covariance index and the water film thickness change index, and the calculation formula is: E d = γ·(C T ·W h + ∥C T - W h ∥) Among them, Ed is the comprehensive environmental disturbance index; γ is the environmental coupling weight coefficient; ∥ ·∥ represents the Manhattan distance between the two indices.

4. The intelligent monitoring system for black titanium surface coloring according to claim 3, wherein: The spectrometer calibration offset rate includes a wavelength baseline offset index; the data sampling period jitter rate includes a timing phase jitter index; the signal baseline drift rate includes a noise covariance index; The calculation formula for the wavelength baseline offset index is: Among them, B λ is the wavelength baseline offset index; λ i is the wavelength deviation of the i-th sampling point; λ 0 is the calibration reference wavelength; δ is the wavelength calibration sensitivity coefficient; The calculation formula for the timing phase jitter index is: Among them, J t is the timing phase jitter index; ti is the time interval of the i-th sampling; is the average sampling period; ∈ is the time jitter correction factor; The calculation formula for the noise covariance index is: Among them, N s is the noise covariance index; S i is the signal baseline drift amount of the i-th sampling; is the mean baseline drift; ζ is the noise suppression coefficient.

5. The intelligent monitoring system for black titanium surface coloring according to claim 4, wherein: The device health index is generated by entropy value fusion of the wavelength baseline offset index, the timing phase jitter index, and the noise covariance index, and the calculation formula is: Among them, D h is the equipment health index; η is the equipment status fusion coefficient.

6. The intelligent monitoring system for black titanium surface coloring according to claim 5, wherein: The wavelet transform decomposition uses the Morlet wavelet basis function to extract multi-scale frequency domain features from the spectral reflectance data, generating a low-frequency approximation component and a high-frequency detail component; each component is input into the coloring quality evaluation model, and the similarity distance with the standard coloring spectrum is calculated.

7. An intelligent monitoring system for black titanium surface coloring according to claim 6, characterized in that: The coloring quality evaluation model is constructed based on the principal component analysis algorithm. By matching the reduced-dimensional spectral feature vectors with a pre-stored standard sample library, the coloring uniformity and the film thickness deviation are determined.

8. An intelligent monitoring system for black titanium surface coloring according to claim 7, characterized in that: It also includes a dynamic calibration module for adaptively adjusting the integration time and gain parameters of the spectrometer according to the environmental disturbance comprehensive index and the equipment health index.

9. The intelligent monitoring system for black titanium surface coloring according to claim 8, characterized in that: The coloring anomaly analysis module uses a support vector machine classifier to perform hyperplane partitioning on the spectral feature vectors, thermal imaging data, and roughness data, identify the types of coloring defects, and output anomaly codes.

Images