Temperature compensation method for data transmission isolation master-slave device communication

Abstract

The application relates to the technical field of industrial digital isolation communication, and discloses a temperature-compensated data transmission isolation master-slave equipment communication method. The method collects the temperature of a master-end digital isolator, the temperature of a slave-end digital isolator and the cable environment temperature, respectively calculates a master-end isolation delay component, a slave-end isolation delay component and a cable propagation delay component according to the three temperature parameters, and synthesizes to obtain a bidirectional link total transmission delay. A baud rate calibration coefficient is calculated based on the bidirectional link total transmission delay, and the baud rate of master-slave equipment communication is synchronously adjusted. The bidirectional delay asymmetry is calculated, and the current baud rate angular frequency is combined to calibrate the receiving end sampling phase, and the master-slave equipment executes data transceiving according to the adjusted parameters. The application solves the half-duplex communication timing mismatch problem caused by the temperature asymmetry of the master and slave ends, and improves the communication reliability in a wide-temperature-range long-distance scene.

Description

Technical Field

[0001] This invention relates to the field of industrial digital isolated communication technology, specifically to a temperature-compensated data transmission isolated master-slave device communication method. Background Technology

[0002] Existing digital isolation master-slave communication systems generally employ single-end temperature compensation or fixed parameter configurations, failing to consider the actual operating conditions where the master and slave devices are located in different ambient temperatures. In industrial settings, the master device is typically located inside a control cabinet, while slave devices are distributed across different locations on the production floor, with temperature differences reaching tens of degrees Celsius. Digital isolators transmit signals via magnetic or capacitive coupling; the carrier mobility of semiconductor devices changes with temperature, causing transmission delay to drift. When the master and slave ends are asymmetrically temperature-dependent, a significant difference exists between the forward and reverse transmission delays, resulting in misalignment of half-duplex direction switching timing. Simultaneously, the propagation speed of electromagnetic waves in twisted-pair cables is affected by the dielectric constant of the medium, which changes slowly with temperature. In long-distance communication scenarios, the coupling effect between cable delay and isolator delay further exacerbates timing deviations. Existing solutions only perform static calibration on one side of the device, failing to compensate for sampling window offsets caused by bidirectional delay asymmetry. This leads to increased communication error rates under high temperature differences, potentially causing communication interruptions in severe cases.

[0003] Therefore, there is an urgent need for a data transmission master-slave communication method that can sense the temperature difference between the two ends and perform dynamic compensation throughout the entire link. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a temperature-compensated data transmission isolation master-slave device communication method, applicable to an isolated communication system comprising a master device and a slave device. The master device is equipped with a master-end digital isolator, and the slave device is equipped with a slave-end digital isolator. The master device and the slave device are connected via a communication cable. The method includes the following steps: S1: Collect the operating temperature of the master digital isolator to obtain the master temperature, collect the operating temperature of the slave digital isolator to obtain the slave temperature, and obtain the cable temperature of the environment in which the communication cable is located. S2: Based on the master end temperature, the slave end temperature, and the cable temperature, calculate the master end isolation delay component, the slave end isolation delay component, and the cable propagation delay component respectively, and synthesize them to obtain the total bidirectional link transmission delay in the master-to-slave direction and the slave-to-master direction; S3: Calculate the baud rate calibration coefficient based on the total transmission delay of the bidirectional link, and synchronously adjust the communication baud rate of the master device and the slave device according to the baud rate calibration coefficient; S4: Calculate the difference between the master-to-slave transmission delay and the slave-to-master transmission delay to obtain the delay asymmetry. Combine this with the angular frequency corresponding to the current baud rate to calculate the optimal sampling phase. Then, calibrate the sampling time of the receiver based on the optimal sampling phase. S5: The master device and the slave device perform data transmission and reception operations according to the adjusted communication baud rate and the calibrated sampling phase.

[0005] Preferably, the steps of acquiring the operating temperature of the master digital isolator to obtain the master temperature and acquiring the operating temperature of the slave digital isolator to obtain the slave temperature include: setting a patch temperature sensor on the master device side, the patch temperature sensor being attached to the housing surface of the master digital isolator to acquire the housing temperature of the master digital isolator; processing the acquired raw temperature data through a first-order inertial filtering algorithm to eliminate high-frequency temperature fluctuations and outputting the master temperature; setting a patch temperature sensor on the slave device side, the patch temperature sensor being attached to the housing surface of the slave digital isolator to acquire the housing temperature of the slave digital isolator; processing the acquired raw temperature data through a first-order inertial filtering algorithm to eliminate high-frequency temperature fluctuations and outputting the slave temperature; when the master device sends a communication frame, it carries the master temperature data in a reserved field in the frame header; when the slave device replies with a response frame, it carries the slave temperature data in a reserved field in the frame header; both parties parse the temperature field in the received frame to complete the bidirectional synchronous exchange of temperature data.

[0006] Further preferably, the step of obtaining the cable temperature of the environment in which the communication cable is located includes: setting temperature acquisition nodes at the beginning, middle, and end of the communication cable, with each temperature acquisition node acquiring the ambient temperature value at the corresponding location; inputting the temperature values ​​at the beginning, middle, and end of the cable into an arithmetic mean calculation unit to calculate the arithmetic mean of the three temperature values; outputting the arithmetic mean as the cable temperature to a delay calculation unit; monitoring the rate of change of the master-end temperature and the slave-end temperature in real time, and immediately initiating a link delay parameter update process when the temperature change at either end exceeds a trigger threshold; and performing delay parameter updates according to a fixed period when the temperature change does not exceed the trigger threshold.

[0007] Further preferably, the step of calculating the baud rate calibration coefficient based on the total transmission delay of the bidirectional link and synchronously adjusting the communication baud rate of the master device and the slave device according to the baud rate calibration coefficient includes: reading a pre-stored reference delay threshold and a reference bit period duration; taking the larger value between the master-to-slave transmission delay and the slave-to-master transmission delay as a representative value of the total transmission delay of the bidirectional link; calculating the difference between the representative value of the total transmission delay of the bidirectional link and the reference delay threshold to obtain the delay increment; dividing the delay increment by the reference bit period duration to obtain the delay percentage change; calculating the baud rate calibration coefficient based on the delay percentage change; multiplying a preset reference baud rate value by the baud rate calibration coefficient to obtain the current optimal working baud rate; the master device generating a baud rate configuration frame, the baud rate configuration frame carrying new baud rate parameters and a switching timestamp; the master device sending the baud rate configuration frame to the slave device, and the slave device returning an acknowledgment response after receiving the configuration frame; and at the time of the switching timestamp, the master device and the slave device simultaneously switching to the new communication baud rate.

[0008] Further preferably, the steps of calculating the master-end isolation delay component, the slave-end isolation delay component, and the cable propagation delay component respectively include: reading the master-end isolator delay parameters, the slave-end isolator delay parameters, and the cable signal propagation speed parameters at the reference temperature; calculating the difference between the master-end temperature and the reference temperature, the difference between the slave-end temperature and the reference temperature, and the difference between the cable temperature and the reference temperature respectively; multiplying the master-end temperature difference by the master-end delay temperature coefficient and then adding it to the reference delay value to obtain the master-end isolation delay component; multiplying the slave-end temperature difference by the slave-end delay temperature coefficient and then adding it to the reference delay value to obtain the slave-end isolation delay component; multiplying the cable temperature difference by the cable speed temperature coefficient and then correcting the reference propagation speed value; and dividing the cable length by the corrected propagation speed value to obtain the cable propagation delay component.

[0009] Further preferably, the step of calculating the difference between the master-to-slave transmission delay and the slave-to-master transmission delay to obtain the delay asymmetry, calculating the optimal sampling phase in combination with the angular frequency corresponding to the current baud rate, and calibrating the sampling time of the receiving end according to the optimal sampling phase includes: subtracting the slave-to-master transmission delay value from the master-to-slave transmission delay value, and using the result as the bidirectional delay asymmetry; multiplying the currently effective baud rate value by twice the constant of pi to obtain the baud rate angular frequency; multiplying the bidirectional delay asymmetry value by the baud rate angular frequency value to obtain the total phase offset; dividing the total phase offset by two to obtain the single-sided compensated phase offset; superimposing the ideal center phase value with the single-sided compensated phase offset, and using the result as the optimal sampling phase; and adjusting the phase offset of the sampling clock according to the optimal sampling phase value so that the sampling trigger time is aligned with the center position of the data bit.

[0010] Further preferably, the method also includes a compensation parameter self-learning optimization step: during communication, communication quality statistics corresponding to different temperature combinations are continuously recorded. The communication quality statistics include the number of error frames, the number of data retransmissions, and the number of successfully received frames. When the accumulated statistics reach the sample size threshold, the parameter optimization process is initiated. The least squares method is used to fit the historical data and optimize and update the isolator temperature coefficient and the cable speed temperature coefficient. The optimized parameters are written to the non-volatile storage unit. Each time the system is powered on, the latest optimized parameters are automatically loaded from the non-volatile storage unit. The parameter self-calibration process is executed according to a fixed time period to compensate for parameter drift caused by device aging and long-term environmental changes.

[0011] More preferably, the steps of the master device and the slave device performing data transmission and reception operations according to the adjusted communication baud rate and the calibrated sampling phase include: adjusting the timing parameters for half-duplex communication direction switching according to the total transmission delay of the bidirectional link; the transmitting end framing and sending data frames according to the adjusted communication baud rate, wherein the data frame sequentially includes a frame start field, an address field, a control field, a data field, and a check field; the receiving end signal is processed by the front-end analog circuit and then input into the digital sampling unit; the digital sampling unit performs level sampling and decision on the received signal according to the calibrated sampling phase; the serial data obtained from the sampling decision is converted from serial to parallel and then input into the verification unit; the verification unit verifies the correctness of the data through the check field, and outputs the data to the upper-layer processing unit after successful verification.

[0012] More preferably, the filtering coefficients of the first-order inertial filtering algorithm are adjusted accordingly based on the temperature change rate; when the temperature change rate is lower than a first rate threshold, the first filtering coefficient is used for processing; when the temperature change rate is higher than a second rate threshold, the second filtering coefficient is used for processing; the first rate threshold is less than the second rate threshold; and the value of the first filtering coefficient is greater than the value of the second filtering coefficient.

[0013] Further preferably, the step of the master device and the slave device simultaneously switching to the new communication baud rate upon reaching the switching timestamp includes: the master device starting a timer after sending a baud rate configuration frame and recording the sending timestamp; the slave device recording the receiving timestamp after receiving the baud rate configuration frame; the slave device calculating the switching time based on the receiving timestamp and link delay parameters to ensure that the time difference between the switching times of the master device and the slave device is less than half a bit period; before the switching time arrives, both parties maintain the original baud rate for communication; when the switching time arrives, both parties simultaneously update the baud rate register configuration; after the configuration update is completed, the master device sends a verification frame, and the slave device receives the verification frame and returns an acknowledgment frame to confirm that the baud rate switching is successful.

[0014] The technical effects include: This invention addresses the timing mismatch issue in half-duplex communication caused by temperature asymmetry between master and slave devices by combining dual-end temperature acquisition with bidirectional link delay synthesis. It dynamically adjusts the baud rate based on the total delay to ensure consistent sampling window margin across different temperatures. The bidirectional delay difference is used to calibrate the received sampling phase, ensuring the sampling point is always aligned with the center of the data eye diagram. This three-level compensation mechanism works synergistically to improve communication reliability in wide-temperature-range, long-distance scenarios. The parameter self-learning function compensates for device aging drift, extending the system's stable operating cycle. Attached Figure Description

[0015] Figure 1 This is a flowchart of the temperature-compensated data transmission isolation master-slave device communication method of the present invention; Figure 2 This is a simulation curve of the temperature transmission delay characteristics of the present invention; Figure 3 This is a simulation diagram of the sampling phase calibration at the receiver end of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0017] Existing technologies suffer from communication timing mismatch when the temperatures of the master and slave ends are asymmetrical.

[0018] Please see Figures 1-3 This solution discloses a temperature-compensated master-slave communication method for data transmission isolation, comprising: acquiring the operating temperature of the master-end digital isolator to obtain the master-end temperature; acquiring the operating temperature of the slave-end digital isolator to obtain the slave-end temperature; and obtaining the cable temperature of the environment in which the communication cable is located. Based on the master-end temperature, the slave-end temperature, and the cable temperature, the master-end isolation delay component, the slave-end isolation delay component, and the cable propagation delay component are calculated respectively, and synthesized to obtain the total bidirectional link transmission delay in the master-to-slave direction and the slave-to-master direction. A baud rate calibration coefficient is calculated based on the total bidirectional link transmission delay, and the communication baud rates of the master device and the slave device are synchronously adjusted according to the baud rate calibration coefficient. The difference between the master-to-slave transmission delay and the slave-to-master transmission delay is calculated to obtain the delay asymmetry. The optimal sampling phase is calculated based on the angular frequency corresponding to the current baud rate, and the sampling time of the receiving end is calibrated according to the optimal sampling phase. The master device and the slave device perform data transmission and reception operations according to the adjusted communication baud rate and the calibrated sampling phase.

[0019] The dual-end temperature acquisition is performed independently on the master and slave sides. A temperature sensing element is mounted on the surface of the master digital isolator, employing a surface-mount package structure and bonded to the isolator housing. The sensing surface of the surface-mount temperature sensor maintains tight contact with the isolator housing, and the contact interface is filled with a thermally conductive medium to reduce thermal resistance. The sensor outputs an analog voltage signal, which is converted into a digital signal by an analog-to-digital converter and then sent to the processing unit. The processing unit filters the raw temperature data to eliminate high-frequency fluctuations introduced by environmental electromagnetic interference. The filtered temperature value is stored in the master device's local storage unit and simultaneously loaded into a reserved field in the frame header each time a communication frame is sent. The slave side uses the same hardware architecture and processing flow. A surface-mount temperature sensor is also mounted on the surface of the slave digital isolator, and the acquired temperature data is filtered and stored in the slave device's local storage unit.

[0020] When the master device initiates communication, the header of the communication frame carries the locally acquired master-end temperature data. After receiving the communication frame, the slave device parses the temperature value in the header field and stores the master-end temperature in its receive buffer. When the slave device generates a reply frame, it similarly loads the locally acquired slave-end temperature data into a reserved field in the header. The master device receives the reply frame and parses it to obtain the slave-end temperature value. Each master-slave data interaction completes one round of bidirectional temperature data synchronization, eliminating the need for a dedicated temperature transmission channel. The temperature data synchronization cycle is consistent with the normal communication cycle, ensuring that the time resolution of the compensation parameter updates matches the data communication frequency.

[0021] The master-end temperature and the slave-end temperature respectively reflect the operating temperature status of the isolation devices at both ends. The digital isolator internally achieves signal transmission across the isolation barrier through magnetic or capacitive coupling structures. The operating temperature of the semiconductor circuits on both sides of the coupling structure directly affects the carrier mobility, thereby altering the signal transmission delay. Collecting the isolator housing temperature can indirectly reflect the junction temperature level of the internal chip. The difference between the housing temperature and the junction temperature is determined by the device's thermal resistance; for a specific type of isolator, the thermal resistance parameter is a fixed value. By extrapolating the junction temperature change trend from the housing temperature, the temperature drift of the isolator's transmission delay can be calculated.

[0022] The cable temperature is acquired through multi-point sampling. The communication cable is laid in an industrial environment, where ambient temperatures may vary at different locations along its length. The cable head, near the main equipment, is typically located inside a control cabinet where the temperature is relatively stable. The middle section of the cable traverses the production area and is significantly affected by ambient temperature fluctuations. The cable tail, near the slave equipment, has a temperature condition dependent on the environment of that slave equipment. Temperature acquisition nodes are set at these three locations, each containing a temperature sensor and signal conditioning circuitry. The temperature values ​​from the three nodes are transmitted to an arithmetic mean calculation unit, which sums the three values ​​and divides by three to obtain the arithmetic mean. This arithmetic mean is then output as the average temperature of the entire cable to the delay calculation unit.

[0023] The temperature change rate monitoring unit tracks the master and slave temperatures in real time. At the end of each sampling period, it calculates the difference between the current temperature and the temperature of the previous period. The absolute value of the difference is divided by the sampling period time to obtain the temperature change rate. The temperature change rate reflects the severity of ambient temperature fluctuations. When the temperature change at either end exceeds the trigger threshold, the link delay parameter update process is immediately initiated to ensure that the compensation parameters quickly follow the temperature change. When the temperature change does not exceed the trigger threshold, the delay parameter is updated at a fixed period to reduce the utilization of computing resources. The trigger threshold value is set according to the system response speed requirements and is determined through a limited number of experiments.

[0024] The total transmission delay of the bidirectional link is calculated using component synthesis. The signal transmission path of the entire communication link sequentially passes through the master-end digital isolator, the communication cable, and the slave-end digital isolator. The delays of these three stages are cascaded and superimposed to form the total transmission delay. The delay of each stage is affected by the temperature at its corresponding location; therefore, a mapping relationship between temperature and delay is established for each stage. The transmission delay of the master-end isolator varies with the master-end temperature, the transmission delay of the slave-end isolator varies with the slave-end temperature, and the propagation delay of the cable varies with the average temperature of the cable. The three delay components are calculated independently and then summed to obtain the total transmission delay of the link.

[0025] The signal transmission path from master to slave is as follows: the signal from the master device first passes through the master-end digital isolator, then propagates through the communication cable, and finally reaches the slave device's receiving end through the slave-end digital isolator. The signal transmission path from slave to master is as follows: the signal from the slave device first passes through the slave-end digital isolator, then propagates through the communication cable, and finally reaches the master device's receiving end through the master-end digital isolator. The cables used in both directions are identical, and the cable delay components are the same. The isolators used in the two directions are in reverse order. The master-end isolation delay and the slave-end isolation delay correspond to different temperature parameters, therefore the total delay values ​​in the two directions may differ. Calculating the total transmission delay in each direction separately yields the bidirectional asymmetric delay parameters.

[0026] The bidirectional delay synthesis of isolated links is calculated using the following formula.

[0027] The formula consists of three terms, corresponding to the master-end isolation delay component, the slave-end isolation delay component, and the cable propagation delay component, respectively. The sum of these three components yields the total transmission delay. (The left side of the formula...) This represents the total transmission delay of the isolated link, measured in seconds, and signifies the total time it takes for a signal to travel from one side of the link to the other. The first item on the right describes the temperature-corrected transmission delay of the master isolator. The transmission delay of the primary isolator at the reference temperature is measured in seconds. It is the nominal delay parameter of the isolator measured under reference temperature conditions. This parameter is obtained from the device manual or actual calibration experiments. The delay temperature coefficient of the main isolator is measured in degrees Celsius. It represents the relative change in delay when the temperature changes by one degree Celsius. This coefficient is determined by the semiconductor material characteristics and circuit structure of the isolator. Different models of devices have different temperature coefficient values. The real-time temperature of the main isolator is measured in degrees Celsius and is obtained by the main temperature acquisition unit. The reference temperature, measured in degrees Celsius, is the reference temperature point for calibrating the delay parameters; it is typically selected from industrial ambient temperature standards. Temperature difference. minus The dimension is Celsius, and the temperature difference is multiplied by the temperature coefficient. Then, multiplying by the degree Celsius value results in dimensionless values, with the result being a dimensionless value. Adding this dimensionless value to the value 1 still yields a dimensionless coefficient, which is then multiplied by the reference delay. The result of multiplying a second by a dimensionless coefficient is still a second. Therefore, the dimension of the entire first term is a second, which is consistent with the dimension on the left side of the formula.

[0028] The mathematical structure of the second term is exactly the same as that of the first term, with each parameter corresponding to a slave isolator. The transmission delay of the slave isolator at the reference temperature is measured in seconds. The delay temperature coefficient of the slave isolator is expressed in degrees Celsius. The real-time temperature of the slave isolator is measured in degrees Celsius. After the same dimensional calculation process, the final result of the second term is also measured in seconds. The master and slave ends can use the same type of isolator. and The values ​​are equal. and The values ​​are equal. Different types of isolation devices can also be used for the master and slave ends. In this case, the two sets of parameters are taken from the calibration values ​​of the corresponding devices.

[0029] The third item describes the signal propagation delay of the cable, among which The effective length of the communication cable, measured in meters, is the actual path length of the signal propagation within the cable. The speed at which a signal propagates in a cable at a reference temperature is measured in meters per second. It is the speed at which electromagnetic waves propagate in the twisted-pair medium, and this speed is determined by the dielectric constant of the cable insulation medium. The temperature coefficient of cable propagation speed, measured in degrees Celsius, represents the relative change in propagation speed when the temperature changes by one degree Celsius. This coefficient is determined by the temperature characteristics of the cable insulation material. This represents the average ambient temperature of the cable, measured in degrees Celsius, and is obtained by averaging temperature data collected from multiple points along the cable. Temperature correction term. The dimensionless quantity is the reference velocity. Even after multiplying by a dimensionless correction factor, the dimension remains meters per second. Cable length Dividing by the corrected propagation speed, the result of dividing meters by meters per second is seconds, therefore the dimension of the third term is seconds.

[0030] All three terms have units of seconds, and the total delay after summing the three terms retains the unit of seconds. The dimensions on both sides of the equation are completely consistent, conforming to the principle of dimensional homogeneity. The theoretical design of this formula is based on semiconductor physics and electromagnetic wave propagation theory. In the semiconductor circuitry inside a digital isolator, carrier mobility decreases with increasing temperature, leading to a longer signal transmission time through the transistor. In industrial temperature ranges, the delay exhibits an approximately linear relationship with temperature; therefore, a first-order linear temperature coefficient model is used to describe the isolator's delay characteristics. The propagation speed of electromagnetic waves in twisted-pair cables depends on the equivalent dielectric constant of the medium. The dielectric constant of polymer insulation materials changes slowly with temperature, causing the propagation speed to drift linearly with temperature. Therefore, the cable delay term uses a propagation time formula corrected for the velocity temperature coefficient.

[0031] The formula derivation starts from the physical essence of signal transmission. The total time a signal takes from the transmitting end of the master device to the receiving end of the slave device is equal to the sum of the transmission times of each segment. For isolators, a reference temperature is defined. The transmission delay is When the temperature changes to At that time, the relative change in delay is directly proportional to the temperature difference, with a proportionality constant of . Therefore, the corrected delay is the base delay multiplied by a temperature coefficient multiplied by the temperature difference. For the cable section, the electromagnetic wave propagation speed... With dielectric constant The propagation speed is inversely proportional to the square root of the dielectric constant. When the dielectric constant changes linearly with temperature, the propagation speed also changes approximately linearly. Therefore, the speed correction form is consistent with the isolator delay correction form. The propagation time equals the cable length divided by the propagation speed; substituting this into the speed correction expression yields the cable delay term. Adding the three delay components gives the total delay formula.

[0032] The baud rate calibration coefficient is calculated based on the ratio of total delay to bit period. In half-duplex communication mode, only one party is allowed to send data at a time, and the direction switching process requires a certain amount of time. The proportion of transmission delay to bit period directly determines the effective margin of the receiver's sampling window. When the temperature increases, leading to an increase in total delay, the proportion of time occupied by direction switching and signal establishment increases accordingly, and the effective sampling window is compressed. If the sampling window is too small, sampling deviation is likely to occur, leading to an increase in the bit error rate. To maintain the sampling margin within the design range, the communication baud rate needs to be reduced proportionally to extend the bit period duration, ensuring that the effective sampling window width for each data bit remains relatively stable.

[0033] When calculating the baud rate calibration coefficient, the larger of the transmission delay in the master-to-slave direction and the transmission delay in the slave-to-master direction is taken as the representative value of the total transmission delay of the bidirectional link. The purpose of selecting the larger value is to compensate based on the worst-case scenario, ensuring that communication in both directions meets the sampling margin requirements. Subtracting the reference delay threshold from the representative value of the total delay yields the delay increment, which reflects the increase in total delay relative to the reference state at the current temperature. Dividing the delay increment by the reference bit period duration yields the change in delay percentage, which represents the proportion of the delay increment to the entire bit period. The change in delay percentage is mapped to the baud rate calibration coefficient through an inverse proportional function; the larger the delay percentage, the smaller the calibration coefficient, and the greater the reduction in baud rate.

[0034] The optimal operating baud rate is obtained by multiplying the preset base baud rate value by the baud rate calibration factor. The base baud rate is the rated communication rate of the system design under normal temperature conditions. When the temperature rises, the calibration factor is less than 1, and the actual baud rate is lower than the base value. When the temperature drops, if the delay is less than the base value, the calibration factor can be greater than 1, and the actual baud rate can be higher than the base value. The baud rate adjustment range is limited by the baud rate range supported by the hardware circuit; when the range is exceeded, the boundary value is taken.

[0035] The master device generates a baud rate configuration frame. This frame includes a command field, a new baud rate parameter field, and a switching timestamp field. The master device sends the baud rate configuration frame to the slave device. Upon receiving the configuration frame, the slave device parses the parameters and returns an acknowledgment frame. After receiving the acknowledgment, the master device waits for the switching time to arrive. At the switching timestamp, both the master and slave devices simultaneously update the baud rate register configuration of their communication interfaces, switching to the new communication baud rate. After the switch is complete, the master device sends a verification frame, and the slave device receives the verification frame and returns an acknowledgment frame, confirming the successful baud rate switch.

[0036] The baud rate adaptive matching coefficient is calculated using the following formula.

[0037] This formula uses an inverse proportional function to calculate the baud rate adjustment factor. The left side of the formula... This represents the baud rate calibration coefficient, a dimensionless value whose range is determined by the delay variation. Multiplying this coefficient by the reference baud rate yields the actual working baud rate. The numerator on the right side of the formula is a pure numerical value of 1, dimensionless. The denominator consists of the numerical value 1 plus a correction term, which describes the degree of influence of the delay variation on the bit period. The correction term... The total transmission delay of the isolation link is expressed in seconds. The larger value of the two-way delay is used in the calculation. The reference delay threshold, measured in seconds, represents the total link delay under normal temperature conditions and serves as the reference delay benchmark during system design. The difference between the two values ​​is... minus The unit of measurement is still seconds. A positive difference indicates an increased delay, while a negative difference indicates a decreased delay.

[0038] The reference bit period is measured in seconds and is calculated by taking the reciprocal of the reference baud rate. The higher the reference baud rate, the shorter the reference bit period. The ratio of the delay difference to the reference bit period is a dimensionless value. This ratio represents the proportion of the delay increment to the entire bit period time. The larger the proportion, the more significant the compression effect of the delay change on the sampling window. This is a delay sensitivity adjustment factor, a dimensionless value, used to control the adjustment range of the baud rate as delay changes. A larger adjustment factor value results in a greater baud rate adjustment for the same delay change; a smaller adjustment factor value results in a smoother baud rate adjustment. The specific value of this factor is set based on the trade-off between system communication reliability and transmission efficiency, and the optimal value is determined through experimental testing.

[0039] Multiplying the delay percentage by the delay sensitivity adjustment factor still yields a dimensionless value. Adding this to the value 1 results in a dimensionless denominator. Dividing the numerator 1 by the dimensionless denominator also yields a dimensionless result, perfectly consistent with the dimensions of the calibration coefficient on the left side of the formula. The design logic of this formula stems from the principle of sampling margin conservation. The ratio of the effective sampling window width to the bit period is defined as the sampling margin coefficient. As the total delay increases, the time occupied by direction switching and signal establishment increases, and the sampling window decreases accordingly. To maintain the sampling margin coefficient constant, the bit period needs to be extended proportionally, and the corresponding baud rate decreases proportionally. The inverse proportional function form enables a smooth mapping relationship of baud rate decrease as delay increases, avoiding communication fluctuations caused by step adjustments.

[0040] The formula derivation begins with the definition of sampling margin. The total delay under the reference state is... The period is The sampling window width is the bit period minus the setup and hold times, both of which are related to the total delay. When the total delay becomes At that time, in order to maintain the sampling margin at the reference level, the bit period A proportional relationship needs to be satisfied. This is achieved by introducing a sensitivity factor. By adjusting the scaling factor, we obtain the relationship between the bit period and the delay. Since the baud rate and bit period are reciprocals, the baud rate calibration factor is the reciprocal of the bit period expansion factor. After algebraic transformation, we obtain the inverse proportional form of the baud rate calibration factor formula described above.

[0041] The sampling phase calibration is performed based on bidirectional delay asymmetry. Under ideal symmetry, the transmission delay in the master-to-slave direction is exactly equal to that in the slave-to-master direction, and the center of the valid data window at the receiver is located at the midpoint of the bit period, corresponding to a phase of π / 2. When the bidirectional delays are unequal, the data arrival time is shifted overall, and the valid data window at the receiver shifts accordingly, resulting in a phase deviation between the window center and the midpoint of the bit period. If sampling is still performed at a fixed phase position, the sampling point will deviate from the window center, reducing the sampling margin, and in severe cases, it may sample data transition edges, leading to bit errors.

[0042] The bidirectional delay asymmetry is obtained by calculating the difference in transmission delay between the two directions. The asymmetry value reflects the degree of difference in delay between the two directions. Multiplying the asymmetry by the angular frequency corresponding to the current baud rate converts the time difference into a phase offset. Since the receiver only needs to compensate for its own phase offset, half of the total phase offset is taken as the single-sided compensation value. The ideal center phase, superimposed with the single-sided compensation value, yields the actual optimal sampling phase. The receiver's digital sampling circuit adjusts the phase offset of the sampling clock according to the optimal sampling phase value, ensuring the sampling trigger time is re-aligned with the center of the data bits, maximizing the sampling window margin.

[0043] The receiver sampling circuit employs a digital delay line structure to achieve phase adjustment. The sampling clock signal passes through multiple delay units, each providing a fixed phase step. By selecting delay paths with different numbers of stages, the phase offset of the sampling clock can be adjusted. The phase adjustment step size is determined by the number of delay units; more stages result in higher phase adjustment resolution, but also higher hardware resource consumption. The optimal sampling phase value is divided by the phase step value to obtain the required number of delay stages. The configuration register selects the corresponding delay path, completing the sampling phase adjustment.

[0044] The dynamic optimization of the received sampling phase is calculated using the following formula.

[0045] The formula consists of two terms: the first is the ideal center phase, and the second is the phase compensation caused by the bidirectional asymmetric delay. The left side of the formula... This indicates the optimal receiving sampling phase, measured in radians, representing the phase position at which the receiver should perform level sampling and decision-making within each data bit cycle. (First item on the right) The first term is a constant equal to half of π, measured in radians, corresponding to the ideal sampling phase at the center of the data bit, i.e., the phase at the quarter-point of the bit period. The second term describes the phase correction caused by the asymmetric delay, where... The bidirectional transmission delay asymmetry is measured in seconds and is calculated by subtracting the transmission delay from the slave to the master direction from the master direction transmission delay. A positive difference indicates that the forward delay is greater than the reverse delay, and a negative difference indicates that the reverse delay is greater than the forward delay.

[0046] This represents the angular frequency corresponding to the current baud rate, measured in radians per second. It is calculated by multiplying the current baud rate by twice pi. Higher baud rates result in higher angular frequencies, leading to greater phase shifts for the same time difference. The time-dimension delay asymmetry is multiplied by the angular frequency, and then multiplied by radians per second, yielding a result in radians. Dividing the total phase shift by 2 still results in radians; this division is necessary because a single receiver only needs to compensate for half of the total shift. The phase correction weighting coefficient is a dimensionless value used to fine-tune the phase compensation intensity based on actual communication performance. The theoretically calculated value may deviate from the actual optimal phase due to factors not included in the model, such as circuit parasitic parameters and trace delays. The weighting coefficient corrects for these deviations. Multiplying the radian value by the dimensionless weighting coefficient preserves the dimension in radians. The sum of the two values ​​gives the overall dimension in radians, perfectly consistent with the phase dimension on the left side of the equation. As an auxiliary unit of measurement for angles, radians are essentially the ratio of arc length to radius, a dimensionless derived quantity. In engineering calculations, the radian dimension is retained to clarify its physical meaning.

[0047] The theoretical basis of this formula stems from the timing relationships of signal transmission. The data signal arrives at the receiver after a link delay, with both the start and end times of the valid data window delayed by the same amount. When there is a phase difference between the reference clocks of the transmitter and receiver, combined with the bidirectional asymmetric delay, the center of the data window observed by the receiver will deviate from the midpoint of the bit period. The time offset corresponding to the delay difference is converted into a phase offset and superimposed on the ideal center phase to obtain the corrected optimal sampling phase. Weighting coefficients are used to compensate for other fixed phase offsets in the actual circuit, ensuring that the calculated result is consistent with the measured optimal sampling point.

[0048] The formula derivation begins with the timing relationship of the data window. Assume the transmitter sends the data at time 0, with a positive delay... The data then arrives at the receiving end of the device, and the starting time of the data window observed by the receiving end is... The data window width is equal to the bit period. Therefore, the window center time is The receiver's local clock operates at bit-period intervals, with the nth sampling time being... To ensure the sampling time is aligned with the center of the window, the sampling time is set equal to the window center time. After phase conversion and algebraic rearrangement, the optimal sampling phase expression is obtained. After introducing weighting coefficients to correct the fixed phase deviation, the above phase optimization formula is obtained.

[0049] The self-learning optimization process for compensation parameters is used to continuously improve the accuracy of the compensation model. During system operation, the communication quality statistics unit continuously records communication status data corresponding to different temperature combinations. The statistical data includes the error frame count, data retransmission count, and the number of successfully received frames. The error frame count records the number of frames that failed verification, the data retransmission count records the number of frames retransmitted due to lack of response, and the number of successfully received frames records the total number of frames that passed verification. These three types of statistical data correspond to different communication quality dimensions, comprehensively reflecting the adaptability of the current compensation parameters.

[0050] When the accumulated statistical data reaches the sample size threshold, the parameter optimization process is initiated. The sample size threshold is set according to the statistical confidence level requirements; a larger sample size results in more reliable fitting results, but also requires more storage and computational resources. The least squares method is used to fit historical data, with the communication quality index as the optimization objective, to recalculate the optimal values ​​for the isolator temperature coefficient and cable speed temperature coefficient. The fitting process uses the current parameters as initial values ​​and iteratively searches for the parameter combination that optimizes the communication quality index.

[0051] The optimized parameters are written to a non-volatile memory unit. This unit is implemented using flash memory or electrically erasable read-only memory, ensuring data retention even when power is off. Each time the system powers on, it automatically loads the latest optimized parameters from the non-volatile memory unit as the initial parameters for compensation calculations. The system performs a parameter self-calibration process at fixed intervals, periodically refitting and updating the parameters to compensate for parameter drift caused by device aging over time. Device aging manifests as a slow change in semiconductor device characteristics over time, with slight shifts in parameters such as the temperature coefficient. Periodic self-calibration can offset the effects of aging, ensuring the compensation accuracy of the system over long-term operation.

[0052] The data transmission and reception process is deeply integrated with the compensation parameter calculation results. The timing parameters for half-duplex communication direction switching are adjusted accordingly based on the total transmission delay of the bidirectional link. The direction switching time is set after a protection period following data transmission, with the protection period length set to a value greater than the total transmission delay, ensuring that the other end has completed data reception when switching directions and avoiding switching conflicts. The sending end assembles and sends data frames according to the currently effective communication baud rate. The data frame sequentially includes a frame start field, an address field, a control field, a data field, and a check field. The frame start field identifies the beginning of a data frame, the address field identifies the target slave device number, the control field distinguishes the frame type, the data field carries the actual transmitted service data, and the check field verifies the correctness of data transmission.

[0053] The input signal at the receiving end is first processed by a front-end analog circuit, which includes an impedance matching network, a filtering circuit, and a level conversion circuit. The impedance matching network ensures that the characteristic impedance of the cable matches the input impedance of the receiver, reducing signal reflection. The filtering circuit removes high-frequency noise components from the signal. The level conversion circuit converts the differential signal into a single-ended logic level. The processed signal is then input to a digital sampling unit. The digital sampling unit samples and determines the level of the received signal according to the calibrated sampling phase; a level higher than the decision threshold at the sampling time is considered high, and a level lower than the decision threshold is considered low.

[0054] The serial data obtained from the sampling decision is converted into parallel data format by the serial-to-parallel conversion unit, and the parallel data is input to the verification unit. The verification unit verifies the correctness of the transmission of the entire frame of data based on the value of the verification field. The verification algorithm adopts cyclic redundancy check. After the verification passes, the data field content is output to the upper-layer processing unit for business logic processing. If the verification fails, the frame of data is discarded, and a retransmission mechanism is triggered according to the protocol rules. The timing parameters of the entire transmission and reception process are driven by the compensation calculation results, realizing closed-loop compensation for temperature drift throughout the entire link.

[0055] The first-order inertial filtering algorithm is used to smooth the raw temperature data. The recursive formula for the first-order inertial filtering is: the current output equals the filter coefficient multiplied by the current input, plus the filter coefficient minus the output of the previous time step. The filter coefficient determines the weight of the new sampled value in the output result. The larger the coefficient, the faster the output follows the input change, and the weaker the smoothing effect; the smaller the coefficient, the smoother the output change, the stronger the smoothing effect, and the slower the response speed.

[0056] The filtering coefficients are adjusted according to the rate of temperature change. When the rate of temperature change is below the first rate threshold, it indicates that the temperature is in a relatively stable state. In this case, the first filtering coefficient is used, and its value is relatively small to enhance the smoothing effect and suppress noise. When the rate of temperature change is above the second rate threshold, it indicates that the temperature is in a rapid change phase. In this case, the second filtering coefficient is used, and its value is relatively large to improve the temperature tracking speed and ensure timely compensation. The first rate threshold is less than the second rate threshold, and there is a transition range between the two thresholds. Within the transition range, the filtering coefficients can be linearly interpolated between the first and second filtering coefficients. This variable coefficient filtering strategy balances steady-state smoothness and dynamic response speed, adapting to different temperature change scenarios.

[0057] The baud rate switching employs a timestamp synchronization mechanism to ensure consistent switching times between the master and slave devices. After sending the baud rate configuration frame, the master device starts its local timing unit to record the timestamp of the frame's transmission. Upon receiving the baud rate configuration frame, the slave device records its receiving timestamp. The slave device calculates the master device's transmission time based on the receiving timestamp and link delay parameters, and then calculates the local switching time using a preset switching delay. The switching time setting ensures that the switching time difference between the master and slave devices is less than half a bit cycle, preventing a baud rate mismatch where one device has switched while the other has not during the switching process.

[0058] Before the handover time arrives, both parties maintain the original baud rate for normal communication. When the handover time arrives, both parties simultaneously update the baud rate register configuration of their communication interfaces. The configuration update takes effect immediately after being written to the register, and the interface circuits send and receive data according to the new baud rate parameters. After the configuration update is complete, the master device sends a verification frame using the new baud rate. The slave device receives the verification frame and performs verification; upon successful verification, it returns an acknowledgment frame. The master device's receipt of the acknowledgment frame confirms successful baud rate switching, and the handover process ends. If no acknowledgment is received within the set time, the master device resends the configuration frame to execute the handover process.

[0059] There are several ways to select a temperature sensor. One approach is to use an integrated digital temperature sensor, which has a built-in analog-to-digital converter and digital interface, directly outputting a digital temperature value; this simplifies the circuit design. Another approach is to use a thermistor paired with an external analog-to-digital converter. Thermistors are less expensive, but require additional signal conditioning circuitry. A thermocouple temperature sensor can also be used; thermocouples are suitable for wide temperature ranges but require cold junction compensation circuitry. All three approaches can achieve the purpose of temperature acquisition; the specific selection depends on the measurement range, accuracy requirements, and cost constraints.

[0060] Digital isolators can be implemented using various technical approaches. Magnetic coupling isolation technology can be used, employing miniature transformers to achieve magnetically coupled signal transmission, resulting in high transmission rates. Capacitive coupling isolation technology can be used, utilizing high-voltage capacitors to achieve electric field coupling transmission, offering strong common-mode transient immunity. Optical coupling isolation technology can be used, employing light-emitting diodes and photosensitive devices for photoelectric conversion; this technology is mature and widely used. All three isolation technologies can achieve both electrical isolation and digital signal transmission; the specific selection depends on the isolation withstand voltage, transmission rate, and interference immunity specifications.

[0061] The type of communication cable can be selected based on the application scenario. Shielded twisted-pair cables can be used; the shielding layer enhances resistance to electromagnetic interference, making them suitable for industrial environments with strong interference. Unshielded twisted-pair cables can be used; they are lower in cost and easier to install, making them suitable for scenarios with weak interference. Coaxial cables can be used; their transmission impedance is stable, making them suitable for high-speed, long-distance transmission. Different cable types have different propagation speeds and temperature coefficients; the corresponding parameters need to be calibrated based on the actual selection.

[0062] The compensation calculation can be performed on both the master and slave devices. One approach is centralized calculation by the master device, where all compensation parameters are calculated, and the results are sent to the slave device via configuration frames for execution. The slave device only configures the parameters. Another approach is distributed calculation, where both the master and slave devices calculate their local compensation parameters, ensuring consistency through a synchronization mechanism. A third approach is slave-based calculation, where all calculations are performed by the slave device, and the master device receives and executes the configuration parameters. All three architectures can achieve temperature compensation; the specific choice depends on the allocation of processor resources and the trade-off between communication overhead.

[0063] The simulation generated two sets of visualizations. The first set shows the characteristic curves of transmission delay versus temperature, with the horizontal axis covering an industrial temperature range of -40°C to 125°C, and the vertical axis representing transmission delay in nanoseconds. The figure plots four curves: the delay component of the master-side digital isolator, the delay component of the slave-side digital isolator, the propagation delay component of the communication cable, and the total bidirectional link delay obtained by combining the three. All delay components increase approximately linearly with temperature, with the reference temperature set at 25°C. The current master-side temperature, slave-side temperature, and cable temperature are marked with vertical lines on the curves for easy observation of the operating point. The information area on the right side of the chart simultaneously displays key parameters calculated based on the patented formula, including the total transmission delay in the master-to-slave direction, bidirectional delay asymmetry, baud rate calibration coefficient, and the adjusted optimal communication baud rate, intuitively demonstrating the impact of temperature changes on link timing and the quantitative basis for baud rate adaptive compensation. The second set of simulations focuses on receiver sampling phase calibration. The screen displays an ideal square wave waveform for three consecutive data bit cycles, with the center of each bit originally the ideal sampling point. When temperature asymmetry causes bidirectional delay differences, the sampling point shifts. The figure shows the actual sampling position after compensation with a solid diamond, and the phase shift amplitude is displayed by dashed lines. A sampling window with the current baud rate bit period as a reference is also plotted, indicating that the sampling time reliably falls within the valid data window after phase calibration. The information bar provides the ideal sampling phase of π / 2, the calculated optimal sampling phase, the corresponding phase shift angle, and the equivalent time shift, verifying that the proposed dynamic sampling phase optimization algorithm can compensate for bidirectional delay asymmetry, ensuring that the sampling trigger edge is always aligned with the center of the data eye diagram, thereby guaranteeing communication reliability in wide-temperature-range, long-distance scenarios.

[0064] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A temperature-compensated digital transmission isolation master-slave device communication method, applied to an isolated communication system including a master device and a slave device, wherein the master device is equipped with a master-end digital isolator, the slave device is equipped with a slave-end digital isolator, and the master device and the slave device are connected via a communication cable, characterized in that, Includes the following steps: S1: Collect the operating temperature of the master digital isolator to obtain the master temperature, collect the operating temperature of the slave digital isolator to obtain the slave temperature, and obtain the cable temperature of the environment in which the communication cable is located. S2: Based on the master end temperature, the slave end temperature, and the cable temperature, calculate the master end isolation delay component, the slave end isolation delay component, and the cable propagation delay component respectively, and synthesize them to obtain the total bidirectional link transmission delay in the master-to-slave direction and the slave-to-master direction; S3: Calculate the baud rate calibration coefficient based on the total transmission delay of the bidirectional link, and synchronously adjust the communication baud rate of the master device and the slave device according to the baud rate calibration coefficient; S4: Calculate the difference between the master-to-slave transmission delay and the slave-to-master transmission delay to obtain the delay asymmetry. Combine this with the angular frequency corresponding to the current baud rate to calculate the optimal sampling phase. Then, calibrate the sampling time of the receiver based on the optimal sampling phase. S5: The master device and the slave device perform data transmission and reception operations according to the adjusted communication baud rate and the calibrated sampling phase.

2. The temperature-compensated data transmission isolation master-slave device communication method according to claim 1, characterized in that, The steps of acquiring the operating temperature of the master digital isolator to obtain the master temperature and acquiring the operating temperature of the slave digital isolator to obtain the slave temperature include: setting a patch temperature sensor on the master device side, the patch temperature sensor being attached to the housing surface of the master digital isolator to acquire the housing temperature of the master digital isolator; processing the acquired raw temperature data through a first-order inertial filtering algorithm to eliminate high-frequency temperature fluctuations before outputting the master temperature; setting a patch temperature sensor on the slave device side, the patch temperature sensor being attached to the housing surface of the slave digital isolator to acquire the housing temperature of the slave digital isolator; processing the acquired raw temperature data through a first-order inertial filtering algorithm to eliminate high-frequency temperature fluctuations before outputting the slave temperature; when the master device sends a communication frame, it carries the master temperature data in a reserved field in the frame header; when the slave device replies with a response frame, it carries the slave temperature data in a reserved field in the frame header; both devices parse the temperature field in the received frame to complete the bidirectional synchronous exchange of temperature data.

3. The temperature-compensated data transmission isolation master-slave device communication method according to claim 1, characterized in that, The step of obtaining the cable temperature of the environment in which the communication cable is located includes: setting temperature acquisition nodes at the beginning, middle, and end of the communication cable, with each temperature acquisition node acquiring the ambient temperature value at the corresponding location; inputting the temperature values ​​at the beginning, middle, and end of the cable into an arithmetic mean calculation unit to calculate the arithmetic mean of the three temperature values; outputting the arithmetic mean as the cable temperature to a delay calculation unit; monitoring the rate of change of the master end temperature and the slave end temperature in real time, and immediately initiating a link delay parameter update process when the temperature change at either end exceeds a trigger threshold; and performing delay parameter updates according to a fixed period when the temperature change does not exceed the trigger threshold.

4. The temperature-compensated data transmission isolation master-slave device communication method according to claim 1, characterized in that, The steps of calculating the baud rate calibration coefficient based on the total transmission delay of the bidirectional link and synchronously adjusting the communication baud rate of the master device and the slave device according to the baud rate calibration coefficient include: reading a pre-stored reference delay threshold and a reference bit period duration; taking the larger value between the master-to-slave transmission delay and the slave-to-master transmission delay as a representative value of the total transmission delay of the bidirectional link; calculating the difference between the representative value of the total transmission delay of the bidirectional link and the reference delay threshold to obtain the delay increment; dividing the delay increment by the reference bit period duration to obtain the delay percentage change; calculating the baud rate calibration coefficient based on the delay percentage change; multiplying a preset reference baud rate value by the baud rate calibration coefficient to obtain the current optimal working baud rate; the master device generating a baud rate configuration frame, which carries new baud rate parameters and a switching timestamp; the master device sending the baud rate configuration frame to the slave device, and the slave device returning an acknowledgment response after receiving the configuration frame; and simultaneously switching to the new communication baud rate at the time of the switching timestamp.

5. The temperature-compensated data transmission isolation master-slave device communication method according to claim 1, characterized in that, The steps of calculating the master-end isolation delay component, slave-end isolation delay component, and cable propagation delay component respectively include: reading the master-end isolator delay parameters, slave-end isolator delay parameters, and cable signal propagation speed parameters at the reference temperature; calculating the difference between the master-end temperature and the reference temperature, the difference between the slave-end temperature and the reference temperature, and the difference between the cable temperature and the reference temperature respectively; multiplying the master-end temperature difference by the master-end delay temperature coefficient and then adding it to the reference delay value to obtain the master-end isolation delay component; multiplying the slave-end temperature difference by the slave-end delay temperature coefficient and then adding it to the reference delay value to obtain the slave-end isolation delay component; multiplying the cable temperature difference by the cable speed temperature coefficient and then correcting the reference propagation speed value; and dividing the cable length by the corrected propagation speed value to obtain the cable propagation delay component.

6. The temperature-compensated data transmission isolation master-slave device communication method according to claim 1, characterized in that, The steps of calculating the difference between the master-to-slave transmission delay and the slave-to-master transmission delay to obtain the delay asymmetry, calculating the optimal sampling phase in combination with the angular frequency corresponding to the current baud rate, and calibrating the sampling time of the receiving end based on the optimal sampling phase include: subtracting the slave-to-master transmission delay value from the master-to-slave transmission delay value, and using the result as the bidirectional delay asymmetry; multiplying the currently effective baud rate value by twice the constant of pi to obtain the baud rate angular frequency; multiplying the bidirectional delay asymmetry value by the baud rate angular frequency value to obtain the total phase offset; dividing the total phase offset by two to obtain the single-sided compensated phase offset; superimposing the ideal center phase value with the single-sided compensated phase offset, and using the result as the optimal sampling phase; and adjusting the phase offset of the sampling clock according to the optimal sampling phase value so that the sampling trigger time is aligned with the center position of the data bit.

7. The temperature-compensated data transmission isolation master-slave device communication method according to claim 1, characterized in that, The method also includes a self-learning optimization step for compensation parameters: during the communication process, communication quality statistics corresponding to different temperature combinations are continuously recorded. The communication quality statistics include the number of error frames, the number of data retransmissions, and the number of frames successfully received. When the accumulated statistics reach the sample size threshold, the parameter optimization process is started. The least squares method is used to fit historical data and optimize and update the isolator temperature coefficient and cable speed temperature coefficient; the optimized parameters are written to the non-volatile memory unit; the latest optimized parameters are automatically loaded from the non-volatile memory unit each time the system is powered on. Perform parameter self-calibration procedures at fixed time intervals to compensate for parameter drift caused by device aging and long-term environmental changes.

8. The temperature-compensated data transmission isolation master-slave device communication method according to claim 1, characterized in that, The steps for the master device and the slave device to perform data transmission and reception operations according to the adjusted communication baud rate and calibrated sampling phase include: timing parameters for half-duplex communication direction switching are adjusted accordingly based on the total transmission delay of the bidirectional link; the transmitting end assembles and sends data frames according to the adjusted communication baud rate, and the data frames sequentially include a frame start field, an address field, a control field, a data field, and a check field; the receiving end signal is processed by the front-end analog circuit and then input into the digital sampling unit; the digital sampling unit performs level sampling and decision on the received signal according to the calibrated sampling phase; the serial data obtained from the sampling decision is converted from serial to parallel and then input into the verification unit; the verification unit verifies the correctness of the data through the check field, and outputs the data to the upper-layer processing unit after successful verification.

9. The temperature-compensated data transmission isolation master-slave device communication method according to claim 2, characterized in that, The filtering coefficients of the first-order inertial filtering algorithm are adjusted according to the temperature change rate; when the temperature change rate is lower than the first rate threshold, the first filtering coefficient is used for processing; when the temperature change rate is higher than the second rate threshold, the second filtering coefficient is used for processing. The first rate threshold is less than the second rate threshold; the value of the first filter coefficient is greater than the value of the second filter coefficient.

10. The temperature-compensated data transmission isolation master-slave device communication method according to claim 4, characterized in that, The step of the master device and the slave device simultaneously switching to the new communication baud rate at the time of the switching timestamp includes: the master device starts timing after sending a baud rate configuration frame and records the sending timestamp; the slave device records the receiving timestamp after receiving the baud rate configuration frame; the slave device calculates the switching time based on the receiving timestamp and the link delay parameter to ensure that the time difference between the switching times of the master device and the slave device is less than half a bit period; before the switching time arrives, both parties maintain the original baud rate for communication; when the switching time arrives, both parties simultaneously update the baud rate register configuration; after the configuration update is completed, the master device sends a verification frame, and the slave device receives the verification frame and returns an acknowledgment frame to confirm that the baud rate switching is successful.

Images