Variable conversion processing method in high-speed rail global communication component manufacturing

Abstract

The application relates to the technical field of signal processing and data conversion in precision measurement, and discloses a variable conversion processing method in high-speed rail global communication component manufacturing, which comprises the following steps: acquiring an original physical variable measurement sequence reflecting the physical state characteristics of an input object; calculating the first-order characteristic component change rate of the sequence at adjacent sampling time points; calculating the second-order difference of the change rate with respect to the sampling time interval to extract a momentum trend factor reflecting the variable acceleration trend characteristics of the sequence; establishing a synergistic mapping rule of the first-order following weight and the second-order prediction gain based on the change rate and the momentum trend factor; generating a frequency pre-activation instruction by using the second-order prediction gain; adjusting the variable conversion processing period according to the instruction; and outputting a target instruction variable. The application captures the momentum trend characteristics of a signal, realizes the advance compensation for the inherent response time delay of a processing link, eliminates phase lag and restrains oscillation.

Description

Technical Field

[0001] This invention belongs to the field of signal processing and data conversion technology in precision measurement, and particularly relates to a variable conversion processing method in the manufacturing of high-speed rail full-domain communication components. Background Technology

[0002] Current mainstream technical solutions employ a fixed execution cycle to perform conversion logic on the collected physical measurement values, converting them into instruction variables required to control the actuators. By maintaining a constant sampling frequency, the timing consistency of the measurement data during transmission is preserved. This approach is widely used in general measurement synchronization scenarios that are not dedicated to specific variables. However, in highly dynamic mobile environments such as high-speed rail, the measured physical variables exhibit extremely rapid change slopes and nonlinear fluctuation characteristics. In mobile environments such as 450 km / h, the evolution rate of physical parameters far exceeds the coverage of conventional sampling frequencies. This causes existing fixed-cycle conversion logic to lag in capturing signal mutation characteristics. This physical delay between the response time and the actual signal measurement time results in phase tearing between the output command and the actual physical state, which is a long-standing, implicit technical debt that is difficult to eliminate in the industry.

[0003] Besides hardware limitations, there are also shortcomings in the control algorithm. For example, Chinese invention patent CN118587896A discloses a real-time calibration method and storage medium for highway traffic operation state simulation data. It uses statistical methods such as Kalman filtering to predict the current state using the previous state and perform residual calibration. Although this method works well for smooth traffic flow data, it exhibits significant phase lag when dealing with variable acceleration signals with high-order derivative characteristics in high-speed rail component manufacturing. Due to the lack of deep capture of signal momentum trends, the algorithm's adjustment speed cannot match the sudden change speed of physical variables, resulting in a time axis offset of the converted command variables, affecting the logical consistency of the precision manufacturing process. To improve these problems, simply increasing the sampling frequency will increase the system's hardware power consumption and data processing pressure, leading to cost bottlenecks in the large-scale application of components. Although the method of adjusting the step size using first-order differential logic can achieve dynamic changes in the step size, it can only perform feedback correction on historical states. When facing non-steady-state conditions with acceleration characteristics, it still cannot overcome the lag window caused by reactive adjustment.

[0004] Therefore, the technical problem to be solved by this invention is how to determine a predictive synchronous conversion mechanism that can identify the momentum of changes in measured values ​​and pre-emptively offset execution delays, thereby improving the tracking accuracy of variable conversion in high dynamic environments while maintaining low power consumption. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: A variable transformation processing method in the manufacturing of high-speed rail full-domain communication components, comprising the following steps:

[0006] Step S1: Real-time acquisition of the original physical variable measurement sequence reflecting the physical state characteristics of the input object. The original physical variable measurement sequence consists of multiple sampling amplitude points arranged continuously on the time domain axis.

[0007] Step S2: Calculate the rate of change of the first-order characteristic components of the original physical variable measurement sequence at adjacent sampling times to characterize the fluctuation slope of the physical state of the input object.

[0008] Step S3: Perform a second difference operation on the rate of change of the first-order characteristic component with respect to the sampling time interval to extract the momentum trend factor that reflects the acceleration trend characteristics of the original physical variable measurement sequence.

[0009] Step S4: Input the first-order characteristic component change rate and momentum trend factor as independent variables into the preset adaptive mapping model. Establish the first-order following weight for adjusting the conversion cycle and the second-order prediction gain for performing advance compensation through the composite mapping function in the adaptive mapping model. Establish the collaborative mapping rule between the first-order following weight, the second-order prediction gain and the variable conversion execution frequency.

[0010] Step S5: Detect the magnitude of the momentum trend factor, and use the second-order prediction gain to generate an execution frequency pre-excitation command before the nonlinear trajectory shift occurs in the original physical variable measurement sequence, so that the phase compensation amount of the execution frequency pre-excitation command on the time axis covers the inherent response delay of the variable transformation processing link.

[0011] Step S6: Based on the frequency adjustment step size contained in the execution frequency pre-excitation command, shorten or extend the execution cycle of variable conversion processing in real time, and output the target command variable after resampling conversion to the command receiving unit so that the target command variable is aligned with the real-time trajectory of the original physical variable measurement sequence in the time domain.

[0012] Preferably, the process of extracting the momentum trend factor in step S3 includes: obtaining the rate of change of the first-order feature component in the current time domain unit and the rate of change of the first-order feature component in the previous adjacent time domain unit; calculating the absolute value of the difference between the rate of change of the first-order feature component in the current time domain unit and the rate of change of the first-order feature component in the previous adjacent time domain unit; and performing a ratio calculation between the absolute value of the difference and a preset basic sampling interval to obtain the momentum trend factor. The calculation formula is as follows: ,in, The rate of change of the first-order characteristic component within the current time-domain unit. Δt represents the rate of change of the first-order characteristic component within the preceding adjacent time-domain unit, and Δt is the basic sampling interval.

[0013] Preferably, the process of establishing the cooperative mapping rule between the first-order following weight and the second-order prediction gain in step S4 includes: determining the base frequency component using the rate of change of the first-order characteristic component, and determining the compensation frequency component using the momentum trend factor; linearly weighting the base frequency component and the compensation frequency component to determine the execution frequency F of the variable transformation process, the calculation formula of which is: , where α is the first-order following weight, with a value ranging from 0.1 to 1.0, and β is the second-order prediction gain, with a value ranging from 0.5 to 2.5.

[0014] Preferably, the process of generating the execution frequency pre-excitation command in step S5 includes: comparing the momentum trend factor with a preset mutation threshold; identifying that the original physical variable measurement sequence has entered a nonlinear jump state when the momentum trend factor exceeds the mutation threshold; increasing the second-order prediction gain in real time according to the magnitude of the momentum trend factor, and determining the frequency increment value corresponding to the execution frequency pre-excitation command based on the increased second-order prediction gain.

[0015] Preferably, the process of outputting the target instruction variable in step S6 includes: obtaining the inherent response delay constant of the variable conversion processing link, wherein the value of the inherent response delay constant is not less than 10ms; performing timestamp shift on the execution frequency pre-excitation instruction according to the inherent response delay constant to generate a timestamp aligned instruction; and using the timestamp aligned instruction to drive the instruction receiving unit.

[0016] Preferably, after step S6, the method further includes: monitoring the real-time computing power occupancy rate of the computing unit; when the momentum trend factor is continuously lower than the preset stability threshold and the computing power occupancy rate exceeds 80%, reducing the first-order following weight α to extend the execution cycle of variable transformation processing.

[0017] Preferably, the mutation threshold is determined by the following steps: acquiring a historical signal time-domain fluctuation data sequence under the measurement environment; extracting historical momentum feature values ​​at trajectory singularities in the historical signal time-domain fluctuation data sequence; calculating the statistical distribution probability of the historical momentum feature values, and setting the value at the 95th percentile of the corresponding probability distribution curve as the mutation threshold.

[0018] Preferably, before calculating the rate of change of the first-order characteristic component in step S2, the method further includes: smoothing the original physical variable measurement sequence using a moving average filtering algorithm to eliminate high-frequency random noise; calculating the amplitude difference between two adjacent sampling times of the smoothed measurement sequence; and performing a ratio calculation between the amplitude difference and a preset basic sampling interval to obtain the rate of change of the first-order characteristic component.

[0019] Preferably, the process of setting the second-order prediction gain β includes: obtaining the upper limit of the rated response frequency of the instruction receiving unit; determining the dynamic compensation range based on the upper limit of the rated response frequency; and performing nonlinear interpolation calculation on the second-order prediction gain using the momentum trend factor within the dynamic compensation range, so that the execution frequency F does not exceed the upper limit of the rated response frequency.

[0020] Preferably, the method further includes: performing continuity verification on the target instruction variable; when the dispersion of the target instruction variable output in two adjacent execution cycles exceeds the preset extreme value of abrupt change, introducing a momentum smoothing operator in the adaptive mapping model, and using the momentum smoothing operator to perform rate limiting filtering on the output step size of the second-order prediction gain in order to correct the phase jump of the target instruction variable.

[0021] Compared with existing technologies, the variable transformation processing method in the manufacturing of high-speed rail full-domain communication components of this invention has the following advantages:

[0022] 1. In the manufacturing of high-speed rail full-domain communication components, by extracting the momentum trend factor that reflects the trend of signal fluctuations and introducing it together with the first-order rate of change into an adaptive mapping model, a predictive synchronous conversion mechanism is constructed. By analyzing the acceleration characteristics of signal changes, in the transient stage when the measured variable enters the variable acceleration range, the execution frequency of the variable conversion logic is adjusted in advance by predicting the compensation component. This frequency pre-excitation action offsets the inherent response delay generated by the physical actuator and the processing link, and achieves high-fidelity alignment of the target physical conversion parameters and the original measurement sequence on the time axis.

[0023] 2. A composite function model combining first-order following weights and second-order prediction gain is adopted to solve the sampling gap problem caused by nonlinear abrupt changes in variables under high dynamic conditions. At the singular point where the trajectory of the measured physical variable undergoes a sharp turn, the momentum trend factor triggers an instantaneous shortening of the execution cycle, enabling the system to pre-set a high-precision sampling window before the slope changes drastically. This processing logic not only eliminates phase tearing caused by response window lag, but also suppresses closed-loop oscillations in the non-steady-state transition process by capturing the momentum characteristics of the signal in real time, ensuring the continuity of the conversion command.

[0024] 3. By using momentum trend factors to identify whether the measured signal is in a steady state or an abrupt change state in real time, the system achieves deep decoupling of variable transformation energy efficiency and dynamic response performance. When the input measurement sequence is in a quasi-steady state, the system extends the processing cycle by reducing the execution frequency, thereby reducing the dynamic switching loss of the computing unit. When the momentum trend factor is detected to exceed the abrupt change threshold, computing resources are precisely allocated to ensure high dynamic tracking accuracy. This avoids the engineering redundancy of blindly increasing the global sampling rate to meet extreme working conditions, and optimizes the global energy efficiency distribution of the global communication components during manufacturing and operation. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the variable transformation process for the momentum trend factor and frequency pre-excitation mechanism of the present invention.

[0026] Figure 2 This is a schematic diagram of the core hardware deployment architecture and signal interaction link of the variable conversion processing system of this invention. Detailed Implementation

[0027] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0028] This invention provides a variable transformation processing method in the manufacturing of high-speed rail full-domain communication components. It comprises an original variable acquisition module, a first-order feature extraction unit, a second-order momentum analysis unit, an adaptive mapping model, and an execution cycle configuration unit. Its core logic lies in capturing the second-order momentum trend of the measurement signal, predicting and adjusting the variable transformation frequency in advance to compensate for the inherent physical delay generated in the processing link. High-speed rail full-domain communication components face complex dynamic environments during manufacturing and operation, especially in high-speed moving scenarios of 450 km / h, where the measured physical variables exhibit nonlinear fluctuation characteristics. To achieve accurate processing of the measured variables… To ensure accurate capture, this invention executes a real-time acquisition procedure for the original physical variable measurement sequence. The original physical variable measurement sequence consists of multiple sampling amplitude points arranged continuously on the time domain axis. The analog signal captured by the sensor is converted into a digital sequence by a high-precision analog-to-digital converter. In this embodiment, the phase component reflecting the physical state characteristics of the input object is selected as the core monitoring indicator, and the basic sampling interval Δt is set to 100μs to ensure that the small fluctuations of the signal are covered at the basic frequency level. By continuously and uninterruptedly acquiring the amplitude of the original sequence, a data foundation with deterministic temporal characteristics is provided for the subsequent slope calculation.

[0029] In the variable transformation process of precision components, relying solely on fixed-frequency sampling will result in lag at the moment of variable abrupt change. After acquiring the measurement sequence, the system performs the calculation of the first-order characteristic component change rate. The characteristic component change rate ΔR is used to characterize the fluctuation slope of the physical state of the input object. The measurement amplitudes at adjacent sampling times are read, the absolute value of their difference is calculated, and it is divided by the basic sampling interval Δt to obtain a scalar value that reflects the first-order fluctuation characteristics of the signal in real time. For example, at time t, if the phase characteristic Φ t It is 1.25 rad, the previous moment If it is 1.22 rad, then the calculated value is... The value is 300 rad / s. This indicator serves as the first reference dimension for adjusting the conversion cycle, achieving an initial match between the sampling rhythm and the signal change slope. Under unsteady conditions such as high-speed trains entering and exiting tunnels or making high-speed curves, the changes in physical parameters exhibit acceleration characteristics, resulting in an unavoidable lag window in traditional reactive regulation. Therefore, the system introduces a momentum trend factor. The extraction procedure is obtained by performing a quadratic difference operation on the rate of change of the first-order feature component with respect to the sampling time interval, that is, calculating the rate of change of the first-order feature component in the current time domain unit. Rate of change of the first-order characteristic component in the preceding adjacent time-domain unit The absolute value of the difference is then used to calculate the ratio between Δt and Δt. The formula is as follows: ,in, Momentum trend factor, The rate of change of the first-order characteristic component within the current time-domain unit. Δt represents the rate of change of the first-order characteristic component within the preceding adjacent time-domain unit, and Δt is the basic sampling interval.

[0030] After receiving the first-order slope and second-order momentum, the system inputs them as independent variables into a preset adaptive mapping model. The model establishes a cooperative mapping rule through its internal composite mapping function to determine the final variable transformation execution frequency F. This mapping model limits the execution frequency F to a binary linear weighted combination of the first-order following weight α and the second-order prediction gain β, calculated as follows: Where F is the variable transformation execution frequency, α is the first-order following weight, and β is the second-order prediction gain. The rate of change of the first-order characteristic component. The momentum trend factor is used. When determining the composite mapping function in the adaptive mapping model, a calibration procedure based on polynomial fitting is executed. A phase simulation signal containing different acceleration amplitudes is input through a signal generator. The phase residual δ of the target command variable relative to the original physical variable measurement sequence is recorded under different combinations of first-order following weight α and second-order prediction gain β. Surface fitting is performed on multiple sets of experimental data using the least squares method to establish the objective function that minimizes the root mean square value of the phase residual δ, and to determine the optimal set of values ​​for α and β. To ensure that the output target command variable maintains timing consistency in the physical link, the execution cycle configuration unit maps the calculated variable transformation processing execution cycle T to the initial value of the hardware timer register loaded by the downstream instruction receiving unit. The system obtains the basic clock frequency of the instruction receiving unit. According to the formula Calculate the timer count value N, where T is the execution cycle of the variable conversion process. With the base clock frequency and N as the timer count value, when the hardware timer count reaches the count value N, an output enable pulse is generated, and the frequency control variable or pointer control variable obtained by resampling in the buffer is written into the physical transmission bus. In this way, the frequency pre-excitation instruction at the logic level is transformed into a hardware trigger action with physical clock precision, eliminating the random phase drift caused by the software scheduling logic at the moment of frequency switching.

[0031] The execution cycle configuration unit adjusts the execution cycle T=1 / F of the variable conversion processing in real time according to the frequency pre-excitation command, and outputs the resampled and converted target command variable to the command receiving unit. This output process includes timestamp shifting for the inherent response delay constant to ensure that the output target physical conversion parameters, namely the frequency control variable used to compensate for signal frequency offset and the pointing control variable used to adjust the spatial transmission direction in real time, remain aligned with the original physical quantities on the time axis. The variable following accuracy of the high-speed rail full-domain communication component remains above 99.5% in extremely dynamic scenarios, and the response time of the conversion logic is controlled within 10ms. Regarding the overlay mechanism of the execution frequency pre-excitation command on the inherent response delay, a first-in-first-out (FIFO) data queue with timestamps is maintained at the command output end to obtain the inherent response delay constant of the variable conversion processing link. After generating the frequency pre-excitation command, according to The numerical value advances the output position of the calculated target instruction variable in the data queue by N basic sampling intervals, where N is... The integer obtained by rounding down the ratio of the ratio to the basic sampling interval Δt is used to compensate for the physical transmission lag caused by the link through phase advance compensation at the logic output time.

[0032] Example 1: In the case of a high-speed train operating at 450 km / h and passing through a large-radius curve, the phase component of the measured physical signal exhibits nonlinear variable acceleration characteristics. Due to the inherent physical delay of 10 ms in the signal acquisition link, data conversion unit, and downstream actuator, the lag window generated by the traditional first-order slope following method at the moment of nonlinear phase change will cause a mismatch between the output target command variable and the original physical variable measurement sequence. To suppress the control system oscillation caused by this physical delay, the system acquires the phase amplitude point in real time with a basic sampling interval Δt of 100 μs. The absolute value of the difference in measurement amplitude between adjacent sampling moments is calculated by the first-order feature extraction unit and divided by Δt to obtain the rate of change of the first-order feature component characterizing the slope of the input object fluctuation. This first-order feature, as an input with deterministic temporal accuracy, directly satisfies the requirements of subsequent second-order momentum analysis to characterize the signal evolution trend.

[0033] To address the variable acceleration characteristics of signal fluctuations, the second-order momentum analysis unit performs a quadratic difference operation on the rate of change of the first-order characteristic components with respect to the sampling time interval, extracting a momentum trend factor that reflects the variable acceleration trend of the sequence. The calculation formula is as follows: ,in, Momentum trend factor, The rate of change of the first-order characteristic component within the current time-domain unit. The first-order characteristic component change rate within the preceding adjacent time-domain cell is given by Δt, which is the basic sampling interval. By introducing this momentum feature, the adaptive mapping model... When the preset mutation threshold is exceeded, a frequency pre-excitation command is generated by increasing the weight of the second-order prediction gain β, thereby offsetting the mismatch between the sampling frequency and the nonlinear change of the signal within a single algorithm logic. The execution cycle configuration unit receives the frequency pre-excitation command and increases the variable conversion execution frequency in advance and shortens the subsequent execution cycle according to the command. This method of adjusting the sampling rhythm in advance by identifying acceleration momentum compensates for the 10ms link delay within the prediction window, reinterpreting the original hysteresis feedback problem as an active frequency compensation method. Ultimately, the frequency control variable used to compensate for the signal frequency offset and the pointing control variable used to adjust the transmission direction in real time are aligned with the original physical quantity on the time axis, so that the variable following accuracy of the train is not less than 99.5% under the dynamic condition of 450km / h, and the response time of the conversion logic is not greater than 10ms.

[0034] Example 2: Experimental verification of the momentum prediction mechanism's effect on offsetting the 10ms physical delay under dynamic conditions of 450km / h. The experimental platform adopted a hardware-in-the-loop simulation system based on solving the dynamic equations. A Doppler phase trajectory data sequence conforming to the motion characteristics of high-speed rail was generated using the Monte Carlo simulation method. This simulation system is equipped with a digital acquisition path with a measurement accuracy of 0.01rad and a sampling frequency upper limit of 500kHz. The core parameter Δt was set to 100μs, and its setting logic is based on the trade-off between the Nyquist sampling criterion and the real-time control step size. When the dynamic bandwidth of the phase signal increases, the upper limit of the sampling frequency is selected to avoid signal aliasing. The settings of the first-order following weight α and the second-order prediction gain β are obtained by balancing the system's steady-state error and overshoot suppression. achieve rad / s Under the operating conditions, by adjusting β, the variable transformation execution frequency F is made to approach the upper limit of its value to compress the response window; in the test environment, additive white Gaussian noise with a signal-to-noise ratio of 20dB is superimposed at the signal source, and a 50Hz power frequency interference harmonic is introduced to form the original physical variable measurement sequence to be processed. One segment of sampling data shows that the original phase amplitude fluctuates between 1.252rad and 1.418rad, accompanied by measurement noise with an amplitude of 0.005rad. This sequence simulates the physical input state of the high-speed rail full-domain communication component in a complex electromagnetic environment. The test group performs differential calculations on the above sequence. Rate of change of the first-order eigencomponent at time t for rad / s, the momentum trend factor is extracted by the second-order momentum analysis unit through quadratic difference. for rad / s At this point, the adaptive image model recognizes that the physical quantity is in the nonlinear acceleration range and triggers the frequency pre-excitation command, which adjusts the variable transformation execution frequency F from 1.0kHz to 2.92kHz.

[0035] The control group used a fixed frequency of 1.0 kHz for variable transformation, resulting in a 12.4 ms time lag at the phase abrupt change in the output target command variable, with a variable following accuracy of 92.1%. In contrast, the experimental group, under the influence of a frequency pre-excitation command, adjusted the sampling window 8.5 ms in advance, aligning the output target command variable with the original physical quantity on the time axis, reducing the lag to 1.2 ms, and achieving a variable following accuracy of 99.7%. Gradient verification was performed by increasing the train speed from 300 km / h to 550 km / h. The experimental results show that with the momentum trend factor... As β increases, the variable transformation execution frequency F exhibits a nonlinear monotonically increasing trend. When β is set to exceed the upper limit of 2.5, the system output exhibits an oscillation with a frequency of 250Hz. This phenomenon conforms to the stability constraint law of the closed-loop system under over-gain state, verifying the rationality of limiting the β range to between 0.5 and 2.5 in this invention. The experimental results confirm that by capturing the second-order momentum characteristics of the signal, the physical delay generated by the processing link can be offset, enabling the high-speed rail full-domain communication component to maintain the phase fidelity of variable transformation under extremely high dynamic environment.

[0036] Example 3: Addressing the technical problem that pressure fluctuations caused by trains passing through tunnels under high-speed operating conditions trigger fine-tuning of the communication component's pointing, leading to nonlinear step characteristics in the phase component, the system implements a sudden change threshold. The offline calibration procedure involves reading the raw physical variable measurement sequence of the sensor in a static state and calculating the standard deviation σ of the second momentum to set a mutation threshold. The standard deviation σ is 3 times, and in this implementation, it takes the value of rad / s This calibration process established the triggering benchmark for the frequency pre-excitation mechanism, providing a basis for quantifying and determining the mutation threshold. Collect no fewer than 10,000 raw sampling amplitude points while the high-speed train is not in operation, and calculate the rate of change of the first-order characteristic component at each sampling time. and momentum trend factor By statistically analyzing the momentum trend factor sequence under static conditions, the standard deviation σ and mean μ of the numerical distribution are extracted, and the mutation threshold is set. It is set to μ+3σ as the boundary between normal fluctuations and nonlinear step states of physical signals.

[0037] The adaptive mapping model receives the momentum trend factor output by the second-order momentum analysis unit. And execute the discrete switching procedure between steady-state mode and transient mode, when momentum trend factor The three consecutive basic sampling intervals Δt are all no greater than the mutation threshold. When the model determines that the physical state is in the linear fluctuation range, the first-order following weight α is set to 1.0 and the second-order prediction gain β is locked at a low gain of 0.5 to maintain the output smoothness of the variable transformation execution frequency F; when the momentum trend factor Exceeding the mutation threshold When the duration reaches two basic sampling intervals Δt, the model triggers a transient pre-excitation mechanism, switching the second-order prediction gain β to the weight cap of 2.2, and calculating the variable transformation execution frequency F according to the formula. Determine the target frequency. In a scenario where the train enters the tunnel at a speed of 450 km / h, when the momentum trend factor... achieve rad / s Subsequently, the variable transformation execution frequency F jumps from 1.0kHz to 3.65kHz within 1.2ms, thereby shortening the response time of the subsequent processing link, compensating for the 10ms physical delay generated between the acquisition module and the instruction receiving unit, and controlling the error envelope of the output target instruction variable after the phase nonlinear change point within 0.008rad, thus achieving the stability of variable transformation in a high dynamic environment.

[0038] Example 4: During the final assembly and commissioning of a new batch of high-speed rail communication components, the system calibrates the inherent response delay constant of the physical link. The signal generator inputs an amplitude value to the original variable acquisition module. The step excitation signal, where, This is the inherent response delay constant of the physical link. The amplitude of the step excitation signal is used, and the logical transition point of the output instruction variable is captured in real time by the instruction receiving unit. The absolute value of the time difference from the excitation application time to the instruction generation time is calculated. Ten sets of data are collected repeatedly, and their arithmetic mean is used as the reference input for timestamp shift compensation. A mapping relationship between the processing link logic depth and the physical clock cycle is established.

[0039] When the system faces parameter drift caused by different manufacturing processes, the adaptive mapping model seeks optimal parameters for the first-order following weight α and the second-order prediction gain β. Under controlled electromagnetic conditions, a dataset containing 10 typical momentum gradients is selected as the input source. The value of β is iteratively adjusted within the model, gradually increasing from 0.5 to 2.5. Here, α is the first-order following weight and β is the second-order prediction gain. The phase residual δ between the output target command variable and the original physical variable measurement sequence is recorded synchronously. Here, δ is the phase residual. The parameter combination that makes the root mean square value of the residual δ tend to the minimum value is selected as the preset weight of the execution cycle configuration unit. After the system completes the initial configuration, it enters the running state, maintaining the stability of the variable transformation logic under different hardware environments.

[0040] Example 5: In the production line debugging environment of high-speed rail full-domain communication components, the system performs the acquisition resolution calibration of the original physical variable measurement sequence. By quantizing the analog signal input in the range of 0 to 5V into a 16-bit digital sequence, the basic bit weight is determined. Under the condition that the basic sampling interval Δt is 100μs, the first-order feature extraction unit calculates the numerical difference between adjacent sampling times. The absolute value of this difference is multiplied by the bit weight to obtain the rate of change of the first-order feature component. The quantization representation, where Δt is the basic sampling interval, The process establishes a benchmark for the transformation of physical state characteristics in the digital domain, using the rate of change of the first-order characteristic component as an example. For the stability adaptation procedure of the adaptive mapping model under different hardware batches, the system selects 1000 sets of phase fluctuation samples simulating a high-speed train operating at 450 km / h to construct an offline validation set. Through iterative calculation, the frequency F for variable transformation is selected to satisfy the formula... The weight combination that minimizes the output phase residual δ is given, where F is the variable transformation execution frequency, α is the first-order follower weight, and β is the second-order prediction gain. Here, δ represents the momentum trend factor, and δ represents the phase residual. When the momentum trend factor output by the second-order momentum analysis unit is detected... When isolated extreme points deviating from physical characteristics are generated by environmental electrostatic pulse interference, the system executes an abnormal data removal procedure by referencing the momentum trend factor in the previous adjacent time domain unit. The current value is used in the calculation, thus maintaining the execution continuity of the processing link for the inherent 10ms response delay compensation mechanism.

[0041] In clustered manufacturing scenarios that support multi-channel parallel processing, the system executes task resource allocation procedures to coordinate variable transformation tasks across different acquisition links, and monitors the rate of change of the first-order feature components corresponding to each channel in real time through the main control module. With momentum trend factor ,in, The rate of change of the first-order characteristic component. The momentum trend factor is used, and the computational load level of each channel is divided into three gradients: high, medium, and low. When a specific channel detects a phase nonlinear step and generates a frequency pre-excitation command, the main control module automatically increases the data bus access priority of that channel and prioritizes the chip select cycle of the floating-point arithmetic unit to shorten its interrupt response time. This priority configuration method based on the dynamic characteristics of variables ensures the real-time performance of key variable conversion tasks in the precision component manufacturing process and avoids secondary processing delays caused by competition for computing resources.

[0042] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. A variable transformation processing method in the manufacturing of high-speed rail full-domain communication components, characterized in that, Includes the following steps: Step S1: Real-time acquisition of the original physical variable measurement sequence reflecting the physical state characteristics of the input object. The original physical variable measurement sequence consists of multiple sampling amplitude points arranged continuously on the time domain axis. Step S2: Calculate the rate of change of the first-order characteristic components of the original physical variable measurement sequence at adjacent sampling times to characterize the fluctuation slope of the physical state of the input object. Step S3: Perform a second difference operation on the rate of change of the first-order characteristic component with respect to the sampling time interval to extract the momentum trend factor that reflects the acceleration trend characteristics of the original physical variable measurement sequence. Step S4: Input the first-order characteristic component change rate and momentum trend factor as independent variables into the preset adaptive mapping model. Establish the first-order following weight for adjusting the conversion cycle and the second-order prediction gain for performing advance compensation through the composite mapping function in the adaptive mapping model. Establish the collaborative mapping rule between the first-order following weight, the second-order prediction gain and the variable conversion execution frequency. Step S5: Detect the magnitude of the momentum trend factor, and use the second-order prediction gain to generate an execution frequency pre-excitation command before the nonlinear trajectory shift occurs in the original physical variable measurement sequence, so that the phase compensation amount of the execution frequency pre-excitation command on the time axis covers the inherent response delay of the variable transformation processing link. Step S6: Based on the frequency adjustment step size contained in the execution frequency pre-excitation command, shorten or extend the execution cycle of variable conversion processing in real time, and output the target command variable after resampling conversion to the command receiving unit so that the target command variable is aligned with the real-time trajectory of the original physical variable measurement sequence in the time domain.

2. The variable transformation processing method in the manufacturing of a high-speed rail full-domain communication component according to claim 1, characterized in that, Step S3 involves extracting the momentum trend factor as follows: obtaining the rate of change of the first-order characteristic component in the current time-domain cell and the rate of change of the first-order characteristic component in the previous adjacent time-domain cell; calculating the absolute value of the difference between the rate of change of the first-order characteristic component in the current time-domain cell and the rate of change of the first-order characteristic component in the previous adjacent time-domain cell; and performing a ratio calculation between the absolute value of the difference and a preset basic sampling interval to obtain the momentum trend factor. The calculation formula is as follows: ,in, The rate of change of the first-order characteristic component within the current time-domain unit. Δt represents the rate of change of the first-order characteristic component within the preceding adjacent time-domain unit, and Δt is the basic sampling interval.

3. The variable transformation processing method in the manufacturing of a high-speed rail full-domain communication component according to claim 1, characterized in that, Step S4, establishing the collaborative mapping rule between the first-order following weights and the second-order prediction gain, includes: determining the base frequency component using the rate of change of the first-order characteristic components, and determining the compensation frequency component using the momentum trend factor; linearly weighting the base frequency component and the compensation frequency component to determine the execution frequency F of the variable transformation process, the calculation formula of which is: , where α is the first-order following weight, with a value ranging from 0.1 to 1.0, and β is the second-order prediction gain, with a value ranging from 0.5 to 2.

5.

4. The variable transformation processing method in the manufacturing of a high-speed rail full-domain communication component according to claim 1, characterized in that, The process of generating the execution frequency pre-excitation command in step S5 includes: comparing the momentum trend factor with a preset mutation threshold; identifying the original physical variable measurement sequence entering a nonlinear jump state when the momentum trend factor exceeds the mutation threshold; increasing the second-order prediction gain in real time according to the magnitude of the momentum trend factor, and determining the frequency increment value corresponding to the execution frequency pre-excitation command based on the increased second-order prediction gain.

5. The variable transformation processing method in the manufacturing of a high-speed rail full-domain communication component according to claim 1, characterized in that, The process of outputting the target instruction variable in step S6 includes: obtaining the inherent response delay constant of the variable conversion processing link; performing time-stamp shifting on the execution frequency pre-excitation instruction according to the inherent response delay constant to generate a time-stamp aligned instruction; and using the time-stamp aligned instruction to drive the instruction receiving unit.

6. The variable transformation processing method in the manufacturing of a high-speed rail full-domain communication component according to claim 1, characterized in that, After step S6, the method further includes: monitoring the real-time computing power occupancy rate of the computing unit; when the momentum trend factor is continuously lower than the preset stability threshold and the computing power occupancy rate exceeds 80%, reducing the first-order following weight α to extend the execution cycle of variable transformation processing.

7. The variable transformation processing method in the manufacturing of a high-speed rail full-domain communication component according to claim 4, characterized in that, The mutation threshold is determined by the following steps: acquiring the historical signal time-domain fluctuation data sequence under the measurement environment; extracting the historical momentum characteristic values ​​at the trajectory singularity points of the historical signal time-domain fluctuation data sequence; calculating the statistical distribution probability of the historical momentum characteristic values, and setting the value at the 95th percentile of the corresponding probability distribution curve as the mutation threshold.

8. The variable transformation processing method in the manufacturing of a high-speed rail full-domain communication component according to claim 1, characterized in that, Before calculating the rate of change of the first-order characteristic component in step S2, the method further includes: smoothing the original physical variable measurement sequence using a moving average filtering algorithm; calculating the amplitude difference of the smoothed measurement sequence in two adjacent sampling times; and performing a ratio calculation between the amplitude difference and a preset basic sampling interval to obtain the rate of change of the first-order characteristic component.

9. The variable transformation processing method in the manufacturing of a high-speed rail full-domain communication component according to claim 3, characterized in that, The process of setting the second-order prediction gain β includes: obtaining the upper limit of the rated response frequency of the command receiving unit; determining the dynamic compensation interval based on the upper limit of the rated response frequency; and performing nonlinear interpolation calculation on the second-order prediction gain within the dynamic compensation interval using the momentum trend factor, so that the execution frequency F does not exceed the upper limit of the rated response frequency.

10. The variable transformation processing method in the manufacturing of a high-speed rail full-domain communication component according to claim 1, characterized in that, The method also includes: performing continuity verification on the target instruction variable; when the dispersion of the target instruction variable output by two adjacent execution cycles exceeds the preset extreme value of abrupt change, introducing a momentum smoothing operator in the adaptive mapping model, and using the momentum smoothing operator to perform rate limiting filtering on the output step size of the second-order prediction gain.

Images