Factory production line data acquisition method and system

Abstract

The invention relates to the technical field of production line data acquisition, and discloses a factory production line data acquisition method and system, and the method comprises the steps: collecting a pressure original signal, an ultrasonic vibration signal and a temperature signal in a target object bonding process in real time, carrying out the mapping of the three signals, generating a pressure quantum state, a vibration quantum state and a temperature quantum state, the target object is a semiconductor production line; by mapping a multi-physical field signal into a quantum state and performing dynamic compensation, complex interference is effectively decoupled, the accuracy of pressure data acquisition is remarkably improved, dynamic coupling characteristics are accurately captured and deviation is corrected through the steps of chaotic entropy extraction, quantum state analysis, unitary operator compensation and the like, and the accuracy of pressure data acquisition is improved. The defect that a conventional calibration method loses efficacy on dynamic working conditions is overcome, so that the collected factory line data better conforms to reality, and the stability of the bonding process is prevented from being judged mistakenly.

Description

Technical Field

[0001] This invention relates to the field of production line data acquisition technology, specifically to a method and system for acquiring data from a factory production line. Background Technology

[0002] Currently, when collecting data from factory production lines, multiple sensors are often used to collect data in order to monitor the status of the production line, provide early warnings of faults, and optimize the production process.

[0003] However, the above-mentioned data acquisition scheme still has significant shortcomings when collecting data from semiconductor production lines. Specifically, semiconductor production lines include many processes, such as bonding, encapsulation, and dicing. In the gold wire bonding process, it is necessary to simultaneously control ultrasonic vibration, bonding pressure, and worktable temperature. During this process, the sensor responsible for monitoring the bonding pressure is simultaneously affected by the mechanical noise of high-frequency vibration and the thermal drift interference of high temperature. These two interference factors are coupled with each other, ultimately causing the collected bonding pressure data to exhibit dynamic distortion. This type of distortion is not due to sensor failure (static calibration results are normal), but rather a coupling effect under dynamic operating conditions. Conventional single-point calibration or offline verification cannot detect it, leading to deviations in subsequent analysis of the collected data and incorrect judgments on the stability of the bonding process. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method and system for data acquisition in factory production lines, thus solving the aforementioned problems.

[0005] The above-mentioned technical objective of the present invention is achieved through the following technical solution: A method for data acquisition from a factory production line, comprising: Step S1: Real-time acquisition of pressure raw signal, ultrasonic vibration signal and temperature signal in the bonding process of the target object, mapping of the three signals to generate pressure quantum state, vibration quantum state and temperature quantum state respectively, the target object is a semiconductor production line; Step S2: Extract the vibrational quantum state and the temperature quantum state to generate a bond entropy flow characteristic value that includes the coupling strength of mechanical noise and thermal drift; Step S3: Analyze the eigenvalues ​​of the bond entropy flow to obtain the dynamic compensation unitary operator acting on the pressure quantum state; Step S4: Transform the pressure quantum state according to the dynamic compensation unitary operator to generate the modified pressure quantum state after interference decoupling; Step S5: Perform quantum measurement on the modified pressure quantum state to generate a true bonding pressure value resistant to dynamic coupling interference.

[0006] Furthermore, the three signals are mapped separately to generate pressure quantum states, vibrational quantum states, and temperature quantum states, including: The original pressure signal, ultrasonic vibration signal, and temperature signal were extracted separately, and the inherent disorder of the captured signal was calculated to obtain three chaotic entropies. The mutual information entropy of the three chaotic entropies is calculated to generate the initial value of the quantum entanglement degree; The three chaotic entropies are converted into qubit sequences respectively, generating three bit strings; Based on the initial value of quantum entanglement, three quasi-quantum states are obtained by calculating the three bit strings; Phase calibration of three quasi-quantum states generates pressure quantum state, vibrational quantum state, and temperature quantum state.

[0007] Furthermore, the original pressure signal, ultrasonic vibration signal, and temperature signal were extracted separately, and the inherent disorder of the captured signals was calculated to obtain three chaotic entropies, including: The original pressure signal, ultrasonic vibration signal, and temperature signal were decomposed to obtain three multi-scale energy percentage sequences. Based on the three multi-scale energy proportion sequences, the dominant scale with the most significant chaotic characteristics of each signal is determined, and three dominant scales are generated. Three fluctuation correlation degrees were obtained by calculating the three dominant scales; Based on the three wave correlation degrees, the effective coverage of the signal chaotic attractor is determined, and the pressure correlation length, vibration correlation length and temperature correlation length are generated. By fusing the pressure correlation length, vibration correlation length, and temperature correlation length, three chaotic entropies are obtained: pressure chaotic entropy, vibration chaotic entropy, and temperature chaotic entropy.

[0008] Furthermore, vibrational quantum states and temperature quantum states are extracted to generate bonded entropy flow eigenvalues ​​that include the coupling strength between mechanical noise and thermal drift, including: Dynamic characteristics of vibrational quantum states and temperature quantum states are analyzed to obtain vibrational quantum transition frequencies and temperature quantum transition frequencies; An interaction matrix is ​​constructed based on the vibrational quantum transition frequency and the temperature quantum transition frequency, thereby generating a vibrational-temperature quantum coupling matrix; Analysis of the vibration-temperature quantum coupling matrix yields the entanglement decay coefficients of the vibrational quantum state and the temperature quantum state. Based on the entanglement decay coefficient, the entropy gradient of vibrational quantum state and temperature quantum state is extracted to obtain the vibration-temperature entropy gradient value; By fusing the vibration-temperature quantum coupling matrix with the entropy gradient value, a bonded entropy flow characteristic value containing the coupling strength of mechanical noise and thermal drift is generated.

[0009] Furthermore, analysis of the eigenvalues ​​of the bond entropy flow yields a dynamic compensation unitary operator acting on the pressure quantum state, including: Temporal analysis is performed on the eigenvalues ​​of the bonded entropy flow to generate the transient coupling coefficients of the entropy flow. Based on the entropy-current transient coupling coefficient, we analyze its interference offset on the pressure quantum state and generate the pressure quantum state interference offset value. The quantum state correction reference value is obtained by calculating the pressure quantum state disturbance offset value; Based on the quantum state correction benchmark, a unitary transformation operator for precise compensation of pressure quantum states is constructed, resulting in a dynamic compensation unitary operator.

[0010] Furthermore, based on the quantum state correction benchmark, a unitary transformation operator for precise compensation of the pressure quantum state is constructed, resulting in a dynamic compensation unitary operator, including: Based on the quantum state correction benchmark value, the dimension and range of pressure quantum state compensation are analyzed, and the benchmark adaptation correction amount is generated; The reference adaptation correction is corrected by combining the real-time characteristics of the disturbance of the pressure quantum state, and the disturbance cancellation calibration is generated. By fusing the reference adaptation correction and the interference cancellation calibration, a dynamic compensation unitary operator is obtained.

[0011] Furthermore, the pressure quantum state is transformed according to the dynamic compensation unitary operator to generate a modified pressure quantum state after disturbance decoupling, including: The dynamic compensation unitary operator and the pressure quantum state are analyzed and dynamically correlated, generating a time-varying coupling transformation weight sequence. Phase amplitude coordinated adjustment of the pressure quantum state is performed based on the time-varying coupled transformation weight sequence to generate a phase amplitude coordinated transformation state; The instantaneous correlation between the phase amplitude co-transformation state and the bond entropy flow eigenvalue is analyzed, the uncancelled mechanical noise-thermal drift coupling interference component is analyzed, and the dynamic coupling residual is generated. Based on the dynamic coupling residual, the phase amplitude cooperative transformation state is compensated to generate the entanglement dimension compensation amount. By fusing the phase amplitude cooperative transformation state with the entanglement dimension compensation quantity, the corrected pressure quantum state after disturbance decoupling is obtained.

[0012] Furthermore, quantum measurements are performed on the modified pressure quantum state to generate a true bonding pressure value resistant to dynamic coupling interference, including: The modified pressure quantum states are screened to identify quantum state components that reflect bonding pressure, thereby generating a sequence of pressure-characteristic quantum observables. By combining the pressure characteristic quantum observable sequence with the bond entropy flow characteristic value, a combination of measurement basis vectors that dynamically adjusts with disturbance is constructed to generate a quantum measurement reference value; Based on the quantum measurement reference value, the corrected pressure quantum state is extracted to generate an instantaneous pressure quantum measurement value vector; Quantum fluctuation processing is applied to the instantaneous pressure quantum measurement value vector to generate a smoothed pressure measurement value; The pressure measurement smoothing value is mapped to generate a true bonding pressure value that is resistant to dynamic coupling interference.

[0013] Furthermore, the smoothed pressure measurement values ​​are mapped to generate true bonding pressure values ​​resistant to dynamic coupling interference, including: Obtain historical pressure data for the bonding process of the target object; By correlating historical pressure data with smoothed pressure measurements, the quantum-physical reference conversion rate was obtained. Based on the real-time changes of the bonding entropy flow characteristic value, the quantum-physical benchmark conversion rate is dynamically calibrated to generate a real-time conversion rate calibration value. Based on the quantum-physics reference conversion rate and the real-time calibration of the conversion rate, the pressure measurement smooth value is calculated to generate a true bonding pressure value that is resistant to dynamic coupling interference.

[0014] Furthermore, a factory production line data acquisition system, applied to the aforementioned factory production line data acquisition method, includes: The acquisition unit is used to acquire the original pressure signal, ultrasonic vibration signal and temperature signal in the bonding process of the target object in real time, and map the three signals to generate pressure quantum state, vibration quantum state and temperature quantum state respectively. The target object is a semiconductor production line. The extraction unit is used to extract vibrational quantum states and temperature quantum states to generate bond entropy flow characteristic values ​​that include mechanical noise and thermal drift coupling strength. The analysis unit is used to analyze the eigenvalues ​​of the bond entropy flow to obtain the dynamic compensation unitary operator acting on the pressure quantum state; The transformation unit is used to transform the pressure quantum state according to the dynamic compensation unitary operator to generate the modified pressure quantum state after interference decoupling; The generation unit is used to perform quantum measurements on the modified pressure quantum state to generate a true bonding pressure value that is resistant to dynamic coupling interference.

[0015] In summary, the present invention has the following main beneficial effects: By converting the raw signals of pressure, vibration, and temperature into quantum states, the limitations of traditional sensor data processing in quantifying dynamic coupling effects are overcome. By extracting the characteristic value of bonding entropy flow, the coupling strength of the two types of interference is accurately captured. Then, a dynamic compensation unitary operator is used to specifically correct the pressure quantum state. Finally, the true pressure value is obtained through quantum measurement. This process achieves optimization from interference source analysis to precise quantum state compensation, effectively eliminating the dynamic distortion problem that cannot be solved by conventional single-point calibration or offline verification. This allows the collected bonding pressure data to truly reflect the actual state of the production line process.

[0016] The actual bonding pressure value can dynamically reflect subtle changes in the process. Based on this data, ultrasonic vibration parameters and table temperature can be adjusted in real time to avoid over-adjustment or under-adjustment caused by pressure data distortion. At the same time, the multi-scale energy ratio analysis and dominant scale screening method in the solution can accurately locate the key frequency band with the most significant signal chaos characteristics, reduce the risk of bonding failure caused by pressure control deviation, and significantly improve the stability of the production line and product yield.

[0017] By transforming the dynamic coupling relationship of physical signals into the entanglement and transition characteristics between quantum states, and using mathematical quantum operators to achieve interference decoupling, a true bonding pressure value is ultimately generated. This approach not only relies on historical data to ensure measurement accuracy but also adapts to dynamic coupling interference in real time through entropy flow changes. Compared to the static limitations of traditional single-point calibration, this dynamic conversion method can accurately eliminate the coupling effects of mechanical noise and thermal drift, ensuring that the collected bonding pressure value truly reflects the bonding process status of the factory line. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the steps of a factory production line data acquisition method according to the present invention; Figure 2 This is a schematic diagram of a factory production line data acquisition system according to the present invention. Detailed Implementation

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] refer to Figure 1 and Figure 2 A method for collecting data from a factory production line, comprising: Step S1: Real-time acquisition of the original pressure signal (bonding pressure signal), ultrasonic vibration signal and temperature signal in the bonding process of the target object, mapping the three signals respectively to generate pressure quantum state, vibration quantum state and temperature quantum state, the target object is a semiconductor production line; Step S2: Extract the vibrational quantum state and the temperature quantum state to generate a bond entropy flow characteristic value that includes the coupling strength of mechanical noise and thermal drift; Step S3: Analyze the eigenvalues ​​of the bond entropy flow to obtain the dynamic compensation unitary operator acting on the pressure quantum state; Step S4: Transform the pressure quantum state according to the dynamic compensation unitary operator to generate the modified pressure quantum state after interference decoupling; Step S5: Perform quantum measurement on the modified pressure quantum state to generate a true bonding pressure value resistant to dynamic coupling interference.

[0021] This solution addresses the issue of pressure data being affected by mechanical noise and thermal drift coupling interference in the bonding process of semiconductor production lines. By using quantum state mapping and entropy flow feature extraction, a compensation operator is dynamically generated to correct the pressure quantum state. The true pressure value is then obtained through quantum measurement. Compared with conventional acquisition methods, this solution can eliminate dynamic coupling interference, avoid data distortion, improve the accuracy of pressure data, facilitate subsequent analysis of process stability, and ensure the efficient and stable operation of the production line.

[0022] In one embodiment, the three signals are mapped to generate a pressure quantum state, a vibrational quantum state, and a temperature quantum state, respectively, including: The original pressure signal, ultrasonic vibration signal, and temperature signal were extracted separately, and the inherent disorder of the captured signal was calculated to obtain three chaotic entropies. The mutual information entropy of the three chaotic entropies is calculated to generate the initial value of quantum entanglement. Specifically, for the three chaotic entropies of pressure, vibration, and temperature, the same number of discrete sampling points are taken to form three sequences. The information entropy of each sequence is calculated: all sampling points are traversed one by one, and the frequency of each value is recorded. The frequency is divided by the total number of sampling points to obtain the probability of each value appearing in the pressure sequence. The probability of each value appearing in the vibration and temperature chaotic entropy sequences is calculated in the same way to obtain the probability of different values ​​appearing in each sequence. The natural logarithm of each probability is taken and multiplied by its own value. All products are summed and the negative value is taken to obtain the result. To obtain the three edge entropies, calculate the joint information entropy of the three chaotic entropies: align the three sequences according to the sampling order, forming a triplet (pressure value, vibration value, temperature value) at each position. Iterate through all positions and record the frequency of each triplet. Divide this frequency by the total number of sampling points to obtain the joint probability of each triplet. Take the natural logarithm of each joint probability and multiply it by its own value. Sum all the products and take the negative value to obtain the joint information entropy. Add the three edge entropies and subtract the joint information entropy to obtain the total mutual information entropy. Divide the total mutual information entropy by the maximum value among the three edge entropies and normalize it to between 0 and 1 to obtain the initial value of the quantum entanglement degree. The three chaotic entropies are converted into qubit sequences to generate three bit strings. Specifically, each sample value of the three chaotic entropies is normalized to [0, 1]. For each normalized value, a threshold of 0.5 is used. If the normalized value is greater than or equal to 0.5, it is encoded as 1. If the normalized value is less than 0.5, it is encoded as 0, forming a single-bit code. All single-bit codes are concatenated in the original sampling order to form the corresponding binary sequence of chaotic entropy, generating three bit strings, namely pressure, vibration, and temperature bit strings. Based on the initial value of quantum entanglement, three quasi-quantum states are obtained by calculating three bit strings. Specifically, for each bit of the three bit strings, it is transformed into a quantum superposition state containing two ground states, |0> and |1>. Analyzing the three bit strings, if the encoding of the current bit is 1, the probability amplitude of the |1> ground state is set to the square root of the initial value of quantum entanglement, and the probability amplitude of the |0> ground state is set to the square root of (1 minus the initial value of quantum entanglement), and the sum of the squares of these two probability amplitudes is 1, forming a superposition state of that bit. If the current bit of the bit string is encoded as 0, the probability amplitude of the ground state |0> is set to the square root of the initial value of quantum entanglement, and the probability amplitude of the ground state |1> is set to the square root of (1 minus the initial value of quantum entanglement). The sum of the squares of these two probability amplitudes is 1, forming a superposition state of the bit. After converting each bit into a superposition state according to this rule, the superposition states at all positions are sequentially connected in the original sampling order of the three bit strings to obtain three quasi-quantum states, namely the pressure quasi-quantum state, the vibration quasi-quantum state, and the temperature quasi-quantum state. Phase calibration is performed on three quasi-quantum states to generate pressure, vibration, and temperature quantum states. Specifically, for each quasi-quantum state, the |0> and |1> probability amplitudes consist of two values: a real part (a value without imaginary units) and an imaginary part (a value with imaginary units). The real part is used as the horizontal length, and the imaginary part as the vertical length. A point is defined on the plane, and a line is drawn from the origin to this point. The angle between this line and the horizontal direction is the phase corresponding to the probability amplitude. Thus, the |0> and |1> phases for pressure, vibration, and temperature can be obtained. For the |0> phase, the three... The |0> phase of each quasi-quantum state is multiplied by the joint information entropy, summed, and then divided by 3 times the joint information entropy. The result is used as the |0> reference phase. The |1> reference phase is calculated in the same way. The difference between the |0> phase of the pressure quasi-quantum state and the |0> reference phase is calculated, and the angle of its |0> probability amplitude is adjusted to make the difference 0. The |0> phases of vibration and temperature are corrected to the |0> reference phase in the same way. This operation is repeated for the |1> phase to make the |1> phases of all three consistent with the |1> reference phase. After traversing all sampling positions and completing the phase correction at each position, the three sequences obtained are the phase-calibrated pressure quantum state, vibration quantum state, and temperature quantum state.

[0023] By extracting the inherent disorder of signals through chaotic entropy and combining it with quantum entanglement to quantify the correlation of multiple signals, traditional sensor data is transformed into quantum state expression. Through phase calibration, the phase characteristics of multi-dimensional signals can be accurately aligned, effectively preserving the coupling relationship of various physical quantities under dynamic conditions. Compared with conventional acquisition methods, it can more realistically reflect the inherent correlation between pressure, vibration and temperature during the bonding process, breaking through the limitation of traditional single-point calibration in failing to capture dynamic coupling effects.

[0024] By using qubit encoding and superposition, the coupling effect of mechanical noise and thermal drift is transformed into a computable quantum state. This quantum state mapping-based processing method can fundamentally distinguish between sensor malfunctions and signal distortion caused by dynamic coupling, thus ensuring the accuracy of data acquisition.

[0025] In one embodiment, the original pressure signal, ultrasonic vibration signal, and temperature signal are extracted separately, and the inherent disorder of the captured signal is calculated to obtain three chaotic entropies, including: The original pressure signal, ultrasonic vibration signal, and temperature signal are decomposed to obtain three multi-scale energy proportion sequences. Specifically, the three signals are divided into multiple sub-signals and a residual signal according to different frequency ranges. The energy of each sub-signal is calculated by squaring the values ​​of all sampling points of the sub-signal and then summing them. The energy of the residual signal is calculated in the same way. The energy of each sub-signal is divided by the total energy (which is the sum of the energies of all sub-signals and the residual signal) to obtain the energy proportion of the corresponding frequency range. The energy of the residual signal is divided by the total energy to obtain the energy proportion of the corresponding frequency range. These energy proportions are arranged in order of frequency range to obtain three multi-scale energy proportion sequences: pressure multi-scale energy proportion sequence, vibration multi-scale energy proportion sequence, and temperature multi-scale energy proportion sequence. Based on three multi-scale energy proportion sequences, the dominant scale with the most significant chaotic characteristics of each signal is determined, and three dominant scales are generated. Specifically, the energy proportion in the three multi-scale energy proportion sequences is used as a probability value. Each probability value is multiplied by the natural logarithm of the energy proportion in the multi-scale energy proportion sequence, and the sum is taken as the negative value to obtain the chaotic entropy of each scale. The chaotic entropy of all scales of the same signal is compared, and the scale with the largest value is taken as the dominant scale with the most significant chaotic characteristics of the signal, thereby generating three dominant scales, namely the pressure-temperature dominant scale, the vibration-temperature dominant scale, and the temperature dominant scale. The calculation of three dominant scales yields three fluctuation correlation degrees. Specifically, for each of the three dominant scales, the corresponding sub-signals are extracted, the average value of all sampling points of each sub-signal is calculated, the average value of each sampling point is subtracted from the average value to obtain the deviation, all deviations are squared and summed, and then divided by the total number of sampling points to obtain the variance. This variance is used as the fluctuation correlation degree, and thus three fluctuation correlation degrees can be obtained, namely pressure fluctuation correlation degree, vibration fluctuation correlation degree, and temperature fluctuation correlation degree. Based on the three fluctuation correlation degrees, the effective coverage of the signal chaotic attractor is determined, and pressure correlation length, vibration correlation length, and temperature correlation length are generated. Specifically, this includes: pairwise combining the values ​​of the sub-signal sampling points of the three dominant scales, and calculating the spatial distance of each pair of points: subtracting the value of the previous sampling point from the value of the subsequent sampling point, and taking the absolute value as the distance between the two points, which is the spatial distance. If the spatial distance is less than the square root of the signal fluctuation correlation degree, it is determined to be a similar point pair; dividing the number of all similar point pairs by the total number of point pairs of the sub-signal (total number of point pairs = total number of points × (total number of points - 1)) yields the correlation integral value corresponding to each spatial distance; using 0.63 times the signal fluctuation correlation degree as a threshold, the correlation integral values ​​are checked in ascending order of distance. When the correlation integral value first reaches or exceeds the threshold... At the threshold, the corresponding distance is the correlation length of the signal, and then the pressure correlation length, vibration correlation length, and temperature correlation length are obtained respectively. Among them, in chaos theory, 1-1 / e (about 0.632) is a typical critical value, where e is the natural constant. When the correlation integral value reaches the critical value, it usually corresponds to the state where the chaotic attractor is effectively covered. That is, the distance at this time can reflect the main spatial scale of the attractor. After exceeding this threshold, the growth of the correlation integral value will slow down significantly. Using 0.63 times the fluctuation correlation degree (variance) as the threshold combines the critical value with the fluctuation range of the signal itself (expressed as variance). This makes the threshold conform to the universal statistical law of chaotic system and adapt to the fluctuation characteristics of signal, ensuring that the correlation length can accurately reflect the actual coverage scale of the chaotic attractor of the signal. By fusing the pressure correlation length, vibration correlation length, and temperature correlation length, three chaotic entropies are obtained: pressure chaotic entropy, vibration chaotic entropy, and temperature chaotic entropy. Specifically, the pressure chaotic entropy is calculated as follows: Pressure chaotic entropy = (pressure correlation length × pressure fluctuation correlation degree + vibration correlation length × vibration fluctuation correlation degree × 0.3 + temperature correlation length × temperature fluctuation correlation degree × 0.3) ÷ (pressure fluctuation correlation degree + vibration fluctuation correlation degree × 0.3 + temperature fluctuation correlation degree × 0.3). Following the above simultaneous calculation method, the vibration chaotic entropy and temperature chaotic entropy can be calculated (each with its own correlation length and fluctuation correlation degree as the main term, and the other two terms multiplied by 0.3).

[0026] By using multi-scale energy decomposition and chaotic entropy calculation, the inherent disorder characteristics of pressure, vibration, and temperature signals can be accurately captured. The signals are divided according to frequency and the energy ratio is extracted. The dominant scale is determined by combining chaos theory. This not only preserves the intrinsic characteristics of each signal, but also quantifies the chaotic characteristics under dynamic working conditions by fusing fluctuation correlation degree and correlation length. This breaks through the limitation of traditional sensor data acquisition that only records surface values. It can deeply mine the dynamic coupling information hidden in the signal, which is convenient for subsequent differentiation of mechanical noise, thermal drift and real pressure signals.

[0027] The generated chaotic entropy highlights the chaotic characteristics of the signal itself while taking into account its correlation with other signals. It effectively characterizes the coupling effect of multiple physical quantities in the bonding process, quantifies and calculates the interference caused by high-frequency vibration and high temperature, which facilitates subsequent quantum state determination. It can accurately identify the pressure data distortion caused by dynamic coupling and improve the reliability of the bonding process state judgment.

[0028] In one embodiment, vibrational quantum states and temperature quantum states are extracted to generate bond entropy flow characteristic values ​​that include the coupling strength between mechanical noise and thermal drift, including: Dynamic characteristics analysis of vibrational quantum states and temperature quantum states yields vibrational quantum transition frequencies and temperature quantum transition frequencies. Specifically, this involves: for vibrational quantum states and temperature quantum states, traversing the ground state codes (0 or 1) of adjacent positions in their sampling sequences. If the codes of adjacent positions are different, it is recorded as a quantum jump. The total number of transitions for each of the vibrational and temperature quantum states is divided by the total number of sampling points of their respective quantum states to obtain the vibrational quantum transition frequencies and temperature quantum transition frequencies. An interaction matrix is ​​constructed based on vibrational quantum transition frequencies and temperature quantum transition frequencies to generate a vibrational-temperature quantum coupling matrix. Specifically, a 2×2 matrix is ​​constructed using vibrational quantum transition frequencies and temperature quantum transition frequencies as basic elements. The elements in the first row and first column of this matrix are vibrational quantum transition frequencies, the elements in the first row and second column and the elements in the second row and first column are the product of the two frequencies, and the elements in the second row and second column are temperature quantum transition frequencies, thereby generating a vibrational-temperature quantum coupling matrix. The vibration-temperature quantum coupling matrix is ​​analyzed to obtain the entanglement decay coefficients of the vibrational quantum state and the temperature quantum state. Specifically, this involves: multiplying the elements of the first row and first column of the vibration-temperature quantum coupling matrix by the elements of the second row and second column, and then subtracting the product of the elements of the first row and second column multiplied by the elements of the second row and first column to obtain the determinant value; adding the elements of the first row and first column of the vibration-temperature quantum coupling matrix to the elements of the second row and second column to obtain the trace value; dividing the determinant value by the trace value and taking its square root to obtain the first value; calculating the arithmetic square root of the product of the vibrational quantum transition frequency and the temperature quantum transition frequency to obtain the second value; and multiplying the first value by the second value to obtain the entanglement decay coefficients of the vibrational quantum state and the temperature quantum state. Based on the entanglement decay coefficient, the entropy change gradients of vibrational quantum states and temperature quantum states are extracted to obtain the vibration-temperature entropy change gradient value. Specifically, this includes: taking the chaotic entropy corresponding to the vibrational quantum state and temperature quantum state as the basis, taking the chaotic entropy of two adjacent points in the sampling order, calculating the difference between the later sampling point and the previous sampling point to obtain the instantaneous entropy change, multiplying the instantaneous entropy change by the entanglement decay coefficient, and then dividing by the sampling time interval between two adjacent points to obtain the entropy change rate, and calculating the average absolute value of the entropy change rate of all sampling points to obtain the vibration-temperature entropy change gradient value. The vibration-temperature quantum coupling matrix is ​​fused with the entropy gradient value to generate a bonded entropy flow characteristic value that includes the coupling strength of mechanical noise and thermal drift. Specifically, the four elements of the vibration-temperature quantum coupling matrix (the vibration quantum transition frequency in the first row and first column, the product of the two frequencies in the first row and second column, the product of the two frequencies in the second row and first column, and the temperature quantum transition frequency in the second row and second column) are multiplied by the vibration-temperature entropy gradient value to obtain four weighted elements. The four weighted elements are arranged in the original matrix positions to form a new matrix. The square values ​​of the four elements in the new matrix are calculated. The square values ​​are added together to obtain the sum. The arithmetic square root of the sum is then taken to obtain the bonded entropy flow characteristic value that includes the coupling strength of mechanical noise and thermal drift.

[0029] By analyzing the transition characteristics of vibrational and temperature quantum states, constructing a coupling matrix, and extracting the entanglement attenuation coefficient, the dynamic coupling strength of mechanical noise and thermal drift is accurately quantified. This allows for the capture of the changing trend of the coupling effect over time. The fused bonding entropy flow feature value can intuitively reflect the coupling degree of the two interferences, facilitating subsequent dynamic compensation of the pressure signal. Furthermore, by deeply fusing quantum transition frequency, coupling matrix, and entropy gradient, the generated bonding entropy flow feature value includes both the static coupling strength of the interference and its dynamic change law, achieving a comprehensive characterization of complex coupling effects. This multi-dimensional feature extraction method breaks through the limitation of traditional sensor data only reflecting a single physical quantity, making the coupling effect of mechanical noise and thermal drift quantifiable and traceable. Consequently, pressure data can be accurately corrected, ensuring the accuracy of the stability analysis of the bonding process.

[0030] In one embodiment, the eigenvalues ​​of the bond entropy flow are analyzed to obtain a dynamic compensation unitary operator acting on the pressure quantum state, including: Temporal analysis of the bonded entropy flow eigenvalues ​​is performed to generate the entropy flow transient coupling coefficient. Specifically, this involves: dividing the time window into fixed 10ms durations; sliding the temporal sequence of the bonded entropy flow eigenvalues ​​point by point within the window; calculating the mean and variance of all bonded entropy flow eigenvalues ​​within each window; taking two consecutive adjacent windows, subtracting the mean of the previous window from the mean of the subsequent window to obtain the mean difference; calculating the variance difference between two consecutive adjacent windows using the same method; dividing the mean difference and variance difference by the 10ms interval between the two windows to obtain the instantaneous rate of change of the mean and the instantaneous rate of change of the variance, respectively; setting the weight of the instantaneous rate of change of the mean to 0.6 and the weight of the instantaneous rate of change of the variance to 0.4; multiplying the instantaneous rate of change of the mean and the instantaneous rate of change of the variance by their respective weights and then summing them to obtain the entropy flow transient coupling coefficient for that time period. Based on the entropy flow transient coupling coefficient, the interference offset caused by it to the pressure quantum state is analyzed, and the pressure quantum state interference offset value is generated. Specifically, within a 10ms time window, the probability amplitudes of |0> and |1> of all sampling points of the pressure quantum state are extracted, the magnitude of each probability amplitude is calculated (the square root of the sum of the squares of the real part and the squares of the imaginary part), and the mean of all magnitudes is calculated. The mean is used as the average magnitude of the pressure quantum state within the window. The average magnitude is multiplied by the entropy flow transient coupling coefficient under the same window to obtain the pressure quantum state interference offset value of the window. The pressure quantum state interference offset value is calculated to obtain the quantum state correction reference value. Specifically, the average value of the pressure quantum state interference offset value of the current window and the previous two windows (a total of three consecutive windows) is calculated, and the average value is normalized to the interval [-1, 1] (by dividing the average value by the absolute value of the maximum value among the three values). The resulting value is the quantum state correction reference value. Based on the quantum state correction benchmark, a unitary transformation operator for precise compensation of pressure quantum states is constructed, resulting in a dynamic compensation unitary operator.

[0031] By analyzing the temporal changes of the bond entropy flow eigenvalues ​​through a fixed 10ms window, and combining the instantaneous rate of change of the mean and variance with weight allocation, the transient coupling coefficient of the entropy flow is accurately generated. This allows for the real-time capture of the instantaneous intensity of dynamic coupling interference. Furthermore, by calculating the correlation between the average modulus of the pressure quantum state and the coupling coefficient, the influence of interference on the shift of the pressure quantum state is quantified, overcoming the limitation of traditional methods in dynamically quantifying interference shifts.

[0032] The correction reference value is obtained by normalizing the mean of the disturbance offset values ​​of three consecutive windows. A dynamic compensation unitary operator is then constructed to achieve accurate and adaptive correction of the pressure quantum state. This correction method based on multi-window historical data can effectively smooth instantaneous disturbance fluctuations and ensure the stability and accuracy of the compensation.

[0033] In one embodiment, based on the quantum state correction reference value, a unitary transformation operator for precise compensation of the pressure quantum state is constructed to obtain a dynamic compensation unitary operator, including: Based on the quantum state correction benchmark value, the dimensions and range of pressure quantum state compensation are analyzed, and the benchmark adaptation correction amount is generated. Specifically, this includes: calculating the phase mean of the pressure quantum states |0> and |1> ground states to obtain two average phases; multiplying the quantum state correction benchmark value by these two average phases respectively to obtain the |0> ground state phase correction amount and the |1> ground state phase correction amount; the two-dimensional vector formed by the |0> ground state phase correction amount and the |1> ground state phase correction amount is the benchmark adaptation correction amount, where the two dimensions correspond to the two compensation dimensions |0> and |1>, and the magnitude of the vector reflects the compensation range. The reference adaptation correction is combined with the real-time characteristics of the pressure quantum state being disturbed to correct the reference adaptation correction and generate the interference cancellation calibration degree. Specifically, the following steps are taken: the pressure quantum state interference offset value of the current window is used as the real-time interference feature; the |0> ground state phase correction and |1> ground state phase correction of the reference adaptation correction are multiplied by the real-time interference feature to obtain two correction components; the arithmetic square root of the sum of squares of the two correction components is calculated to obtain the interference cancellation calibration degree. The reference adaptation correction and the interference cancellation calibration are fused to obtain the dynamic compensation unitary operator. Specifically, this includes: calculating the product of the ground state phase corrections of |0> and |1> in the reference adaptation correction with the interference cancellation calibration, and then adding these two products to obtain the total phase offset angle; using the total phase offset angle as the rotation angle, constructing a matrix according to the rotation angle: filling the first cell of the first row and the second cell of the second row with the cosine value of half the rotation angle; filling the second cell of the first row with the negative imaginary unit (-i) multiplied by the sine value of half the rotation angle; filling the first cell of the second row with the imaginary unit (i) multiplied by the sine value of half the rotation angle. This matrix is ​​the dynamic compensation unitary operator.

[0034] By analyzing the dimensions and range of pressure quantum state compensation, a benchmark adaptation correction is generated. Combined with real-time interference feature correction, the interference cancellation calibration degree is obtained. Finally, a dynamic compensation unitary operator is constructed. Through a two-dimensional phase correction method, the interference shift of the |0> and |1> ground states can be accurately covered. The unitary operator in the form of a rotation matrix can achieve directional correction while keeping the quantum state mode length unchanged. Compared with the shortcomings of conventional methods that cannot specifically cancel dynamic coupling interference, this scheme can adapt to the coupling changes of mechanical noise and thermal drift in real time, providing accurate dynamic compensation for pressure quantum states and effectively eliminating signal distortion caused by coupling effects.

[0035] In one embodiment, the pressure quantum state is transformed according to the dynamic compensation unitary operator to generate a modified pressure quantum state after interference decoupling, including: The dynamic compensation unitary operator and the pressure quantum state are analyzed and dynamically correlated to generate a time-varying coupling transformation weight sequence. Specifically, the modulus of the probability amplitudes of |0> and |1> in the pressure quantum state are multiplied by the elements of the first row and first column and the second row and second column of the dynamic compensation unitary operator, respectively, to obtain two correlation values. The two correlation values ​​are added together, and the result is the coupling transformation weight at the current time. The coupling transformation weight at each time is calculated sequentially according to the sampling time order. The coupling transformation weights at all times are arranged to form the time-varying coupling transformation weight sequence. Phase and amplitude coordinated adjustment of the pressure quantum state is performed based on the time-varying coupled transformation weight sequence to generate a phase and amplitude coordinated transformation state. Specifically, this includes: multiplying the weight of each sampling time in the time-varying coupled transformation weight sequence with the phase of the ground state of the pressure quantum state at the corresponding time |0> and |1>, respectively, to obtain the adjusted phases of the ground state of |0> and |1>, respectively; multiplying the magnitude of the probability amplitude of the ground state of |0> and |1> at the sampling time by the weight in the time-varying coupled transformation weight sequence at the corresponding sampling time, respectively, to obtain the adjusted amplitudes of the ground state of |0> and |1>, respectively; for the ground state of |0>... For the ground state |1>, multiply the adjusted amplitude by the cosine of the adjusted phase to obtain the new real part; multiply the adjusted amplitude by the sine of the adjusted phase to obtain the new imaginary part; thus, obtain the new real and imaginary parts of the ground state |0> and the ground state |1> respectively. The new real and imaginary parts constitute the new probability amplitude, which is the new probability amplitude of the ground state |0> and the ground state |1> respectively. Combine the new probability amplitudes of |0> and |1> at each time step to obtain the quantum state at that time step. Connect the quantum states at all time steps in the order of sampling to obtain the phase-amplitude cooperative transformation state. The instantaneous correlation between the phase amplitude co-transformation state and the bond entropy flow eigenvalue is analyzed, and the uncancelled mechanical noise-thermal drift coupling interference component is analyzed to generate a dynamic coupling residual. Specifically, this includes: for each sampling time, calculating the sum of the magnitudes of the probability amplitudes of |0> and |1> in the phase amplitude co-transformation state at that sampling time, and taking it as the instantaneous intensity of the transformation state; calculating the absolute difference between the instantaneous intensity of the transformation state and the bond entropy flow eigenvalue at the same moment, and then calculating the average of the absolute differences at all times. This average value is the dynamic coupling residual, which is used to reflect the uncancelled coupling interference component. Based on the dynamic coupling residual, the phase amplitude cooperative transformation state is compensated to generate the entanglement dimension compensation amount. Specifically, this includes multiplying the magnitudes of the probability amplitudes of |0> and |1> in the phase amplitude cooperative transformation state, multiplying the product by the dynamic coupling residual, and then dividing by the initial value of the quantum entanglement degree to obtain the entanglement dimension compensation amount. The phase amplitude cooperative transform state and the entanglement dimension compensation amount are fused to obtain the corrected pressure quantum state after interference decoupling. Specifically, this involves: for each sampling time, multiplying the real part of the new probability amplitude of the |0> ground state in the phase amplitude cooperative transform state by the entanglement dimension compensation amount at that sampling time, and using the product as the real part of the corrected |0> ground state probability amplitude; then multiplying the imaginary part of the new probability amplitude of the |0> ground state by the entanglement dimension compensation amount at that time, and using the product as the imaginary part of the corrected |0> ground state probability amplitude; for the |1> ground state, repeating the same processing steps as the |0> ground state, that is, multiplying the real part and imaginary part of its new probability amplitude by the entanglement dimension compensation amount at that sampling time to obtain the real part and imaginary part of the corrected |1> ground state probability amplitude; combining the two corrected ground state probability amplitudes at each time according to the sampling order to form the quantum state at that time, and then arranging the quantum states at all times in sequence to obtain the corrected pressure quantum state after interference decoupling.

[0036] The phase and amplitude of the pressure quantum state are coordinated by dynamically coupling and transforming the weight sequence. The dynamic compensation unitary operator and the characteristics of the pressure quantum state are accurately correlated to ensure that the correction process conforms to the characteristics of real-time interference. Through the calculation of dynamic coupling residual degree, the uncancelled coupling interference can be effectively identified. Then, a second correction is performed by the entanglement dimension compensation amount to form a closed loop optimization, thereby reducing the coupling effect of mechanical noise and thermal drift and improving the purity of the pressure quantum state.

[0037] By integrating the phase amplitude co-transformation state and the entanglement dimension compensation, the generated modified pressure quantum state can maximally remove dynamic coupling interference. Through time-by-time fine correction, it not only retains the intrinsic characteristics of the pressure signal, but also specifically eliminates the distortion caused by coupling, which helps to accurately obtain the true bonding pressure value and facilitates subsequent analysis of the stability of the bonding process.

[0038] In one embodiment, quantum measurement is performed on the modified pressure quantum state to generate a true bonding pressure value resistant to dynamic coupling interference, including: The modified pressure quantum states are screened to identify quantum state components reflecting bonding pressure, generating a sequence of pressure characteristic quantum observables. Specifically, this involves: for each modified pressure quantum state, iterating through the ground state probability amplitudes (|0> and |1>) at each sampling time, calculating the squared magnitudes (the sum of the squared real and imaginary parts) of both, and simultaneously calculating the mean of the original pressure signal. Subtracting the mean from the original signal value at each time moment yields the pressure deviation value. For each time moment, the squared magnitudes of |0> and |1> are compared with the corresponding pressure deviation value using Pearson correlation coefficient calculations. Specifically, when calculating the Pearson correlation coefficient between the squared magnitude sequence and the pressure deviation value sequence, the average values ​​of all values ​​in the squared magnitude sequence and the average values ​​in the pressure deviation value sequence are first calculated. For each time moment, the average value of the squared magnitude sequence is subtracted from the squared magnitude at that time moment to obtain the deviation of the squared magnitude at that time moment. Finally, the average value of the pressure deviation value sequence is subtracted from the pressure deviation value at that time moment to obtain the pressure deviation at that time moment. The deviation of the pressure deviation value is calculated as follows: Multiply the two deviations at each time step, sum the products at all times, and divide by the total number of samples minus one to obtain the covariance of the two deviations; Squar the deviations of |0> modulus squares at each time step, sum them, divide by the total number of samples minus one, and take the square root to obtain the standard deviation of the |0> modulus squares sequence; Squar the deviations of the pressure deviation values ​​at each time step, sum them, divide by the total number of samples minus one, and take the square root to obtain the standard deviation of the pressure deviation value sequence; Divide the previously obtained covariance by the product of these two standard deviations. The Pearson correlation coefficient between the |0> modulus square sequence and the pressure deviation value sequence can be obtained. Following the same calculation method, replacing the |0> modulus square sequence with the |1> modulus square sequence yields the Pearson correlation coefficient between the |1> modulus square sequence and the pressure deviation value sequence. Quantum state components (including the corresponding |0> and |1> modulus squares) with an absolute Pearson correlation coefficient greater than or equal to 0.8 are selected. These quantum state components are then integrated according to the sampling order to generate a pressure characteristic quantum observable sequence. By combining the pressure characteristic quantum observable sequence with the bond entropy flow characteristic value, a combination of measurement basis vectors that dynamically adjusts with disturbance is constructed to generate a quantum measurement reference value. Specifically, this includes: normalizing the bond entropy flow characteristic value sequence to [0, 1] as a dynamic weight; squaring the magnitudes of |0> and |1> at each time step in the pressure characteristic quantum observable sequence, multiplying them by the dynamic weights at the corresponding time steps, and then adding the two to obtain the measurement basis vector value at that time step; arranging the measurement basis vector values ​​at all times in the sampling order; assigning the same weight to each basis vector value in the sequence (the sum of all weights is 1); and multiplying each basis vector value by its corresponding weight and then summing the results to obtain the quantum measurement reference value. Based on the quantum measurement reference value, the modified pressure quantum state is extracted to generate an instantaneous pressure quantum measurement value vector. Specifically, for the modified pressure quantum state, the square of the magnitude of the ground state probability amplitude of each time step |0> and |1> is multiplied by the quantum measurement reference value, and the two products are added together to obtain the instantaneous pressure quantum measurement value at that time step. The instantaneous pressure quantum measurement value at each time step is calculated sequentially according to the sampling order. The instantaneous pressure quantum measurement values ​​at all times are arranged sequentially, and the ordered array formed is the instantaneous pressure quantum measurement value vector. Quantum fluctuation processing is applied to the instantaneous pressure quantum measurement vector to generate a smoothed pressure measurement value. Specifically, the instantaneous pressure quantum measurement vector is divided into sliding windows based on three consecutive time points. For each window, the sum of the instantaneous pressure quantum measurement values ​​at the three time points is calculated and divided by 3 to obtain the window mean. The measurement value at the middle time point of the window is replaced with this mean. If the absolute difference between the measurement value at the first and last time points of the window and the mean exceeds 10% of the mean, the values ​​at these two time points are replaced with the mean. After sliding processing window by window, a correction sequence with the same length as the original vector is formed. The sum of all values ​​in the correction sequence is then calculated and divided by the sequence length to obtain the smoothed pressure measurement value. The pressure measurement smoothing value is mapped to generate a true bonding pressure value that is resistant to dynamic coupling interference.

[0039] By using the Pearson correlation coefficient to screen out quantum state components that are strongly correlated with pressure deviation, the characteristic information reflecting the real bonding pressure is accurately extracted, irrelevant interference components are eliminated, and a dynamic measurement basis vector is constructed by combining the bonding entropy flow characteristic value. This allows the measurement reference to be adaptively adjusted according to the intensity of interference, breaking through the limitation of traditional fixed references in dealing with dynamic coupling. This targeted screening and dynamic adaptation method improves the targeting of quantum measurement.

[0040] By processing instantaneous quantum measurement values ​​through a sliding window, fluctuations caused by quantum fluctuations are effectively smoothed out. At the same time, outliers deviating from the mean are corrected to ensure the stability of the smoothed pressure measurement values. The final mapped true bonding pressure value completely eliminates the coupling interference of mechanical noise and thermal drift, solving the problem of pressure data distortion under dynamic operating conditions. Compared with conventional acquisition methods, the data acquired by this solution can better reflect the true state of the bonding process and reduce misjudgments caused by data deviations.

[0041] In one embodiment, mapping the smoothed pressure measurement values ​​to generate a true bonding pressure value resistant to dynamic coupling interference includes: Obtain historical pressure data for the bonding process of the target object; the historical pressure data is the historical pressure data for bonding. The quantum-physical reference conversion rate is obtained by correlating historical pressure data with pressure measurement smoothing values. Specifically, this includes: extracting samples from historical pressure data that are consistent with the current process conditions; pairing the historical actual pressure value and the pressure measurement smoothing value at the corresponding time; calculating the ratio of the historical actual pressure value to the pressure measurement smoothing value in all data pairs; and calculating the average of these ratios, which is the quantum-physical reference conversion rate. Based on the real-time changes in the characteristic values ​​of bond entropy flow, the quantum-physical benchmark conversion rate is dynamically calibrated to generate a real-time calibration value for the conversion rate. Specifically, this involves: extracting the real-time sequence of the characteristic values ​​of bond entropy flow in 10ms windows; calculating the ratio of the difference between the mean values ​​of adjacent windows to the mean value of the previous window; using this ratio as the real-time rate of change of entropy flow; calculating the mean of the real-time rate of change of entropy flow in the current window and the two windows before it; multiplying the mean value by a weight of 0.2 and then multiplying it by the quantum-physical benchmark conversion rate to obtain the real-time calibration value for the conversion rate. Based on the quantum-physical reference conversion rate and the real-time conversion rate calibration value, the pressure measurement smooth value is calculated to generate the true bonding pressure value that is resistant to dynamic coupling interference. Specifically, the quantum-physical reference conversion rate and the real-time conversion rate calibration value are added to obtain the total conversion rate. The pressure measurement smooth value is multiplied by the total conversion rate to obtain the true bonding pressure value that is resistant to dynamic coupling interference.

[0042] By correlating historical pressure data with pressure measurement smoothing values ​​to determine the baseline conversion rate, and combining the real-time changes in bonding entropy flow characteristic values ​​to dynamically calibrate the conversion rate, the true bonding pressure value is finally generated. This approach relies on historical data to ensure measurement accuracy and can adapt to dynamic coupling interference in real time through entropy flow changes. Compared with the static limitations of traditional single-point calibration, the dynamic conversion method of this solution can accurately eliminate the coupling effects of mechanical noise and thermal drift, so that the output pressure value truly reflects the bonding process status.

[0043] In one embodiment, a factory production line data acquisition system, applied to the aforementioned factory production line data acquisition method, includes: The acquisition unit is used to acquire the original pressure signal, ultrasonic vibration signal and temperature signal in the bonding process of the target object in real time, and map the three signals to generate pressure quantum state, vibration quantum state and temperature quantum state respectively. The target object is a semiconductor production line. The extraction unit is used to extract vibrational quantum states and temperature quantum states to generate bond entropy flow characteristic values ​​that include mechanical noise and thermal drift coupling strength. The analysis unit is used to analyze the eigenvalues ​​of the bond entropy flow to obtain the dynamic compensation unitary operator acting on the pressure quantum state; The transformation unit is used to transform the pressure quantum state according to the dynamic compensation unitary operator to generate the modified pressure quantum state after interference decoupling; The generation unit is used to perform quantum measurements on the modified pressure quantum state to generate a true bonding pressure value that is resistant to dynamic coupling interference.

[0044] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for data acquisition in a factory production line, characterized in that, include: Step S1: Real-time acquisition of pressure raw signal, ultrasonic vibration signal and temperature signal in the bonding process of the target object, mapping of the three signals to generate pressure quantum state, vibration quantum state and temperature quantum state respectively, the target object is a semiconductor production line; Step S2: Extract the vibrational quantum state and the temperature quantum state to generate a bond entropy flow characteristic value that includes the coupling strength of mechanical noise and thermal drift; Step S3: Analyze the eigenvalues ​​of the bond entropy flow to obtain the dynamic compensation unitary operator acting on the pressure quantum state; Step S4: Transform the pressure quantum state according to the dynamic compensation unitary operator to generate the modified pressure quantum state after interference decoupling; Step S5: Perform quantum measurement on the modified pressure quantum state to generate a true bonding pressure value resistant to dynamic coupling interference.

2. The factory production line data acquisition method according to claim 1, characterized in that, The three signals are mapped separately to generate pressure quantum state, vibrational quantum state, and temperature quantum state, including: The original pressure signal, ultrasonic vibration signal, and temperature signal were extracted separately, and the inherent disorder of the captured signal was calculated to obtain three chaotic entropies. The mutual information entropy of the three chaotic entropies is calculated to generate the initial value of the quantum entanglement degree; The three chaotic entropies are converted into qubit sequences respectively, generating three bit strings; Based on the initial value of quantum entanglement, three quasi-quantum states are obtained by calculating the three bit strings; Phase calibration of three quasi-quantum states generates pressure quantum state, vibrational quantum state, and temperature quantum state.

3. The factory production line data acquisition method according to claim 2, characterized in that, The original pressure signal, ultrasonic vibration signal, and temperature signal are extracted separately, and the inherent disorder of the captured signals is calculated to obtain three chaotic entropies, including: The original pressure signal, ultrasonic vibration signal, and temperature signal were decomposed to obtain three multi-scale energy percentage sequences. Based on the three multi-scale energy proportion sequences, the dominant scale with the most significant chaotic characteristics of each signal is determined, and three dominant scales are generated. Three fluctuation correlation degrees were obtained by calculating the three dominant scales; Based on the three wave correlation degrees, the effective coverage of the signal chaotic attractor is determined, and the pressure correlation length, vibration correlation length and temperature correlation length are generated. By fusing the pressure correlation length, vibration correlation length, and temperature correlation length, three chaotic entropies are obtained: pressure chaotic entropy, vibration chaotic entropy, and temperature chaotic entropy.

4. The factory production line data acquisition method according to claim 2, characterized in that, Vibrational and temperature quantum states are extracted to generate bonded entropy flow eigenvalues ​​that include the coupling strength between mechanical noise and thermal drift, including: Dynamic characteristics of vibrational quantum states and temperature quantum states are analyzed to obtain vibrational quantum transition frequencies and temperature quantum transition frequencies; An interaction matrix is ​​constructed based on the vibrational quantum transition frequency and the temperature quantum transition frequency, thereby generating a vibrational-temperature quantum coupling matrix; Analysis of the vibration-temperature quantum coupling matrix yields the entanglement decay coefficients of the vibrational quantum state and the temperature quantum state. Based on the entanglement decay coefficient, the entropy gradient of vibrational quantum state and temperature quantum state is extracted to obtain the vibration-temperature entropy gradient value; By fusing the vibration-temperature quantum coupling matrix with the entropy gradient value, a bonded entropy flow characteristic value containing the coupling strength of mechanical noise and thermal drift is generated.

5. The factory production line data acquisition method according to claim 4, characterized in that, Analysis of the eigenvalues ​​of the bond entropy flow yields a dynamic compensation unitary operator acting on the pressure quantum state, including: Temporal analysis is performed on the eigenvalues ​​of the bonded entropy flow to generate the transient coupling coefficients of the entropy flow. Based on the entropy-current transient coupling coefficient, we analyze its interference offset on the pressure quantum state and generate the pressure quantum state interference offset value. The quantum state correction reference value is obtained by calculating the pressure quantum state disturbance offset value; Based on the quantum state correction benchmark, a unitary transformation operator for precise compensation of pressure quantum states is constructed, resulting in a dynamic compensation unitary operator.

6. The factory production line data acquisition method according to claim 5, characterized in that, Based on the quantum state correction reference value, a unitary transformation operator for precise compensation of the pressure quantum state is constructed, resulting in a dynamic compensation unitary operator, including: Based on the quantum state correction benchmark value, the dimension and range of pressure quantum state compensation are analyzed, and the benchmark adaptation correction amount is generated; The reference adaptation correction is corrected by combining the real-time characteristics of the disturbance of the pressure quantum state, and the disturbance cancellation calibration is generated. By fusing the reference adaptation correction and the interference cancellation calibration, a dynamic compensation unitary operator is obtained.

7. A method for acquiring data from a factory production line according to claim 6, characterized in that, The pressure quantum state is transformed using a dynamic compensation unitary operator to generate a modified pressure quantum state after interference decoupling, including: The dynamic compensation unitary operator and the pressure quantum state are analyzed and dynamically correlated, generating a time-varying coupling transformation weight sequence. Phase amplitude coordinated adjustment of the pressure quantum state is performed based on the time-varying coupled transformation weight sequence to generate a phase amplitude coordinated transformation state; The instantaneous correlation between the phase amplitude co-transformation state and the bond entropy flow eigenvalue is analyzed, the uncancelled mechanical noise-thermal drift coupling interference component is analyzed, and the dynamic coupling residual is generated. Based on the dynamic coupling residual, the phase amplitude cooperative transformation state is compensated to generate the entanglement dimension compensation amount. By fusing the phase amplitude cooperative transformation state with the entanglement dimension compensation quantity, the corrected pressure quantum state after disturbance decoupling is obtained.

8. A method for acquiring data from a factory production line according to claim 7, characterized in that, Quantum measurements are performed on the modified pressure quantum state to generate a true bonding pressure value resistant to dynamic coupling interference, including: The modified pressure quantum states are screened to identify quantum state components that reflect bonding pressure, thereby generating a sequence of pressure-characteristic quantum observables. By combining the pressure characteristic quantum observable sequence with the bond entropy flow characteristic value, a combination of measurement basis vectors that dynamically adjusts with disturbance is constructed to generate a quantum measurement reference value; Based on the quantum measurement reference value, the corrected pressure quantum state is extracted to generate an instantaneous pressure quantum measurement value vector; Quantum fluctuation processing is applied to the instantaneous pressure quantum measurement value vector to generate a smoothed pressure measurement value; The pressure measurement smoothing value is mapped to generate a true bonding pressure value that is resistant to dynamic coupling interference.

9. A method for acquiring data from a factory production line according to claim 8, characterized in that, Mapping the smoothed pressure measurement values ​​to generate true bond pressure values ​​resistant to dynamic coupling interference includes: Obtain historical pressure data for the bonding process of the target object; By correlating historical pressure data with smoothed pressure measurements, the quantum-physical reference conversion rate was obtained. Based on the real-time changes of the bonding entropy flow characteristic value, the quantum-physical benchmark conversion rate is dynamically calibrated to generate a real-time conversion rate calibration value. Based on the quantum-physics reference conversion rate and the real-time calibration of the conversion rate, the pressure measurement smooth value is calculated to generate a true bonding pressure value that is resistant to dynamic coupling interference.

10. A factory production line data acquisition system, applied to the factory production line data acquisition method according to any one of claims 1-9, characterized in that, include: The acquisition unit is used to acquire the original pressure signal, ultrasonic vibration signal and temperature signal in the bonding process of the target object in real time, and map the three signals to generate pressure quantum state, vibration quantum state and temperature quantum state respectively. The target object is a semiconductor production line. The extraction unit is used to extract vibrational quantum states and temperature quantum states to generate bond entropy flow characteristic values ​​that include mechanical noise and thermal drift coupling strength. The analysis unit is used to analyze the eigenvalues ​​of the bond entropy flow to obtain the dynamic compensation unitary operator acting on the pressure quantum state; The transformation unit is used to transform the pressure quantum state according to the dynamic compensation unitary operator to generate the modified pressure quantum state after interference decoupling; The generation unit is used to perform quantum measurements on the modified pressure quantum state to generate a true bonding pressure value that is resistant to dynamic coupling interference.

Images