A program control method for illumination regulation

Abstract

The present application relates to the technical field of industrial control system, disclose a kind of program control method for illumination regulation, comprising: program control unit obtains illumination sampling sequence from illumination detection channel and is smoothed as steady-state illumination data, the first-order difference value of adjacent period steady-state illumination data is calculated to determine illumination time-varying gradient absolute value;When absolute value is less than or equal to mutation threshold constant, based on standard proportional integral gain coefficient closed loop output control signal;When absolute value is greater than mutation threshold constant, disconnect proportional integral loop and suspend feedback operator, extract historical control signal mean as static constant write light intensity power regulation unit, the present application reconstructs path timing, decouples conventional feedback and transient disturbance, effectively eliminates overshoot oscillation of light source transition section.

Description

Technical Field

[0001] This invention belongs to the field of industrial control system technology, and in particular relates to a program control method for illuminance regulation. Background Technology

[0002] Currently, illuminance regulation in industrial automation typically employs a closed-loop control method that combines program control with continuous feedback. The program control unit calculates the pulse width modulation signal based on the real-time illuminance residual between the real-time collected illuminance data and the set target value, and adjusts the output power of the light source through the light intensity power adjustment unit to stabilize the illuminance in the process environment near the set target value. The continuous feedback control loop continuously calculates and accumulates the error according to a fixed feedback gain, so that the illuminance gradually tends to stabilize.

[0003] In semiconductor lithography workshops or high-precision optical inspection environments, frequent shuttling of automated material conveyors can cause short-term strong shading, resulting in abrupt changes in illuminance. The dynamic influence between multiple light sources can also introduce high-frequency additional interference. When these external interferences are superimposed, the program segment switching and real-time error feedback in traditional control systems operate simultaneously, and illuminance detection data is prone to sudden changes. The proportional-integral calculation loop may treat this short-term step disturbance as a continuous steady-state deviation, thereby outputting excessively large adjustment commands, causing large fluctuations in light source power. For example, Chinese invention patent application CN102914945A discloses a distributed exposure dose control system and method. This scheme uses independent calibration models for each sub-unit for feedforward and feedback compensation. However, this distributed independent calibration model relies excessively on static calibration tables and parameter fitting curves. When facing complex nonlinear dynamic conditions or thermal drift caused by long-term operation, the dynamic coupling between sub-units is difficult to completely decouple by the static independent model, resulting in lag and residual errors in the compensation factor, which cannot meet the stringent requirements of high-precision lithography for dynamic dose extreme value control.

[0004] Therefore, how to determine the absolute value of the time-varying illuminance gradient based on the difference between the steady-state illuminance data of the current sampling period and the previous sampling period, and when the absolute value of the time-varying illuminance gradient is greater than the constant threshold of sudden change, to activate the second control flow based on the characterization state, to disconnect the proportional-integral calculation loop within the transient time window, and at the same time to call the average value of the control signal of the historical steady-state operation period as the control output quantity and directly write it into the light intensity power adjustment unit, is a technical problem that existing illuminance program control needs to solve. Summary of the Invention

[0005] To address the problems in the background art, the technical solution of the present invention is as follows: A program control method for illuminance regulation, comprising the following steps:

[0006] Step S1: The program control unit acquires the illuminance sampling sequence from the illuminance detection channel at a fixed sampling period, smooths the illuminance sampling sequence using a moving average filtering algorithm, and outputs the steady-state illuminance data for the current sampling period.

[0007] Step S2: The program control unit calculates the first-order difference between the steady-state illuminance data of the current sampling period and the steady-state illuminance data of the adjacent previous sampling period to determine the absolute value of the time-varying gradient of illuminance.

[0008] Step S3: The program control unit compares the absolute value of the time-varying illuminance gradient with a preset mutation threshold constant to determine the control shunting state: when the absolute value of the time-varying illuminance gradient is less than or equal to the mutation threshold constant, the characterization state is established as the first value; when the absolute value of the time-varying illuminance gradient is greater than the mutation threshold constant, the characterization state is established as the second value.

[0009] Step S4: The program control unit allocates a control path based on the characterization state: when the characterization state is the first value, the first control flow is activated, and the real-time illuminance residual between the steady-state illuminance data of the current sampling period and the set target value is calculated based on the standard proportional-integral gain coefficient, and the control signal is output to the light intensity power adjustment unit; when the characterization state is the second value, the second control flow is activated, the proportional-integral calculation loop is disconnected and the feedback operator is suspended, and the average value of the historical control signal is extracted as a static constant within the set transient time window, and the register value of the light intensity power adjustment unit is rewritten through the static constant.

[0010] Preferably, the sampling period is 10ms; in the moving average filtering algorithm, the program control unit extracts the illuminance sample values ​​of 5 consecutive sampling periods including the current sampling period, calculates the arithmetic mean of the illuminance sample values ​​of 5 sampling periods, and establishes the arithmetic mean as the steady-state illuminance data of the current sampling period; the mutation threshold constant is 15lx / ms; the width of the transient time window is 3 sampling periods, corresponding to a time length of 30ms; the average of the historical control signals is the average of the control signals output to the light intensity power adjustment unit in the 10 consecutive steady-state operation periods before the current sampling period.

[0011] Preferably, when the number of consecutive operating cycles with the first value representing the state reaches the set steady-state counting threshold, the path for enabling the first control flow in step S4 further includes the following sub-steps: Step S41, the program control unit obtains the temperature sensor reading from the temperature sensor of the current environment and smoothly converts it into the temperature drift parameter of the current environment; Step S42, the program control unit makes a slight correction to the standard proportional-integral gain coefficient based on the temperature drift parameter, limiting the correction amplitude to within ±2% of the initial value of the standard proportional-integral gain coefficient, so that the steady-state illuminance data is stabilized within the error range corresponding to the set target value.

[0012] Preferably, step S4 includes the following sub-steps during the period when the proportional-integral calculation loop of the second control flow is disconnected: Step S43, the independent safety detection module obtains the absolute illuminance value from the illuminance detection channel at a control cycle of 10ms and continuously monitors the absolute illuminance value to determine whether the absolute illuminance value deviates from the set target value; Step S44, when the absolute illuminance value deviates from the set target value by more than the set safety boundary threshold, the program control unit forcibly interrupts the signal output of the second control flow and resets the control signal sent to the light intensity power adjustment unit to the set 50% fixed duty cycle safety signal.

[0013] Preferably, the path for enabling the first control flow in step S4 includes the following sub-steps: Step S45, the program control unit subtracts the set target value from the steady-state illuminance data of the current sampling period to obtain the current real-time illuminance residual; Step S46, the program control unit multiplies the real-time illuminance residual by the proportional coefficient and the integral coefficient respectively to obtain the proportional term and the integral term, and combines the proportional term and the integral term to generate a control signal for driving the light intensity power adjustment unit.

[0014] Preferably, the path for extracting the historical control signal mean in step S4 includes the following sub-steps: Step S47, when the representation state is the first value, the program control unit continuously writes the control signal output in each sampling period into the buffer, maintaining a dynamic update queue of control signals containing 10 consecutive steady-state cycles; Step S48, when the representation state switches to the second value, the program control unit reads all control signals in the dynamic update queue, calculates the arithmetic mean of all control signals, and establishes the arithmetic mean as a static constant.

[0015] Preferably, the steady-state counting threshold is 500 consecutive operating cycles.

[0016] Preferably, the safety boundary threshold is 5% of the set target value.

[0017] Preferably, the control signal received by the light intensity power adjustment unit is a square wave pulse width modulation signal, and the duty cycle adjustment range of the pulse width modulation signal is limited to a closed interval of 1% to 99%.

[0018] Compared with existing technologies, the program control method for illuminance regulation of the present invention has the following advantages:

[0019] 1. In the program control of illuminance regulation, normal illuminance changes and transient step disturbances are distinguished based on the absolute value of the time-varying gradient of illuminance. The first control flow and the second control flow are activated respectively. When a sudden change in illuminance occurs, the proportional-integral calculation loop is temporarily disconnected, and the power output of the light source is maintained by the average value of the control signal of the historical steady-state operating cycle. This avoids the proportional-integral calculation loop from treating short-term disturbances as steady-state deviations and continuously adjusting. Therefore, program segment switching or external pulse interference will not cause over-adjustment, which can reduce overshoot and oscillation in the transition stage of the light source, while taking into account both the accuracy of steady-state illuminance control and the stability of transient response.

[0020] 2. When the system is running in a continuous steady state, the present invention makes a small correction to the standard proportional-integral gain coefficient based on the temperature drift parameter, which can compensate for the temperature drift caused by the continuous operation of the light-emitting device and reduce the accumulation of illuminance control deviation after long-term operation. This processing does not rely on complex multivariate mapping calculations, has a small amount of computation, is suitable for industrial controllers with limited computing power, and is conducive to maintaining the long-term stable operation of the illuminance adjustment loop.

[0021] 3. During the operation of the second control flow, the present invention continuously monitors the absolute illuminance value by an independent safety detection module. When the illuminance deviates from the set target value by more than the safety boundary threshold, the second control flow can be interrupted in time, and the control signal can be reset to a fixed duty cycle safety signal. Thus, even if a strong interference occurs again during the temporary disconnection of the proportional-integral calculation loop, the light source output can be limited, and the control loop can be prevented from losing its constraint due to the continuous use of historical control signals, thereby improving the safety of illuminance regulation under sudden operating conditions. Attached Figure Description

[0022] Figure 1 This is a flowchart of the illumination sampling filtering and control path splitting process of the present invention;

[0023] Figure 2 This is the multi-level control flow and safety interruption state transition diagram of the present invention. Detailed Implementation

[0024] 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.

[0025] A programmable control method for illuminance regulation includes the following steps:

[0026] Step S1: The program control unit acquires the illuminance sampling sequence from the illuminance detection channel at a fixed sampling period, smooths the illuminance sampling sequence using a moving average filtering algorithm, and outputs the steady-state illuminance data for the current sampling period.

[0027] Step S2: The program control unit calculates the first-order difference between the steady-state illuminance data of the current sampling period and the steady-state illuminance data of the adjacent previous sampling period to determine the absolute value of the time-varying gradient of illuminance.

[0028] Step S3: The program control unit compares the absolute value of the time-varying illuminance gradient with a preset mutation threshold constant to determine the control shunting state: when the absolute value of the time-varying illuminance gradient is less than or equal to the mutation threshold constant, the characterization state is established as the first value; when the absolute value of the time-varying illuminance gradient is greater than the mutation threshold constant, the characterization state is established as the second value.

[0029] Step S4: The program control unit allocates a control path based on the characterization state: when the characterization state is the first value, the first control flow is activated, and the real-time illuminance residual between the steady-state illuminance data of the current sampling period and the set target value is calculated based on the standard proportional-integral gain coefficient, and the control signal is output to the light intensity power adjustment unit; when the characterization state is the second value, the second control flow is activated, the proportional-integral calculation loop is disconnected and the feedback operator is suspended, and the average value of the historical control signal is extracted as a static constant within the set transient time window, and the register value of the light intensity power adjustment unit is rewritten through the static constant.

[0030] Preferably, the sampling period is 10ms; in the moving average filtering algorithm, the program control unit extracts the illuminance sample values ​​of 5 consecutive sampling periods including the current sampling period, calculates the arithmetic mean of the illuminance sample values ​​of 5 sampling periods, and establishes the arithmetic mean as the steady-state illuminance data of the current sampling period; the mutation threshold constant is 15lx / ms; the width of the transient time window is 3 sampling periods, corresponding to a time length of 30ms; the average of the historical control signals is the average of the control signals output to the light intensity power adjustment unit in the 10 consecutive steady-state operation periods before the current sampling period.

[0031] Preferably, when the number of consecutive operating cycles with the first value representing the state reaches the set steady-state counting threshold, the path for enabling the first control flow in step S4 further includes the following sub-steps: Step S41, the program control unit obtains the temperature sensor reading from the temperature sensor of the current environment and smoothly converts it into the temperature drift parameter of the current environment; Step S42, the program control unit makes a slight correction to the standard proportional-integral gain coefficient based on the temperature drift parameter, limiting the correction amplitude to within ±2% of the initial value of the standard proportional-integral gain coefficient, so that the steady-state illuminance data is stabilized within the error range corresponding to the set target value.

[0032] Preferably, step S4 includes the following sub-steps during the period when the proportional-integral calculation loop of the second control flow is disconnected: Step S43, the independent safety detection module obtains the absolute illuminance value from the illuminance detection channel at a control cycle of 10ms and continuously monitors the absolute illuminance value to determine whether the absolute illuminance value deviates from the set target value; Step S44, when the absolute illuminance value deviates from the set target value by more than the set safety boundary threshold, the program control unit forcibly interrupts the signal output of the second control flow and resets the control signal sent to the light intensity power adjustment unit to the set 50% fixed duty cycle safety signal.

[0033] Preferably, the path for enabling the first control flow in step S4 includes the following sub-steps: Step S45, the program control unit subtracts the set target value from the steady-state illuminance data of the current sampling period to obtain the current real-time illuminance residual; Step S46, the program control unit multiplies the real-time illuminance residual by the proportional coefficient and the integral coefficient respectively to obtain the proportional term and the integral term, and combines the proportional term and the integral term to generate a control signal for driving the light intensity power adjustment unit.

[0034] Preferably, the path for extracting the historical control signal mean in step S4 includes the following sub-steps: Step S47, when the representation state is the first value, the program control unit continuously writes the control signal output in each sampling period into the buffer, maintaining a dynamic update queue of control signals containing 10 consecutive steady-state cycles; Step S48, when the representation state switches to the second value, the program control unit reads all control signals in the dynamic update queue, calculates the arithmetic mean of all control signals, and establishes the arithmetic mean as a static constant.

[0035] Preferably, the steady-state counting threshold is 500 consecutive operating cycles.

[0036] Preferably, the safety boundary threshold is 5% of the set target value.

[0037] Preferably, the control signal received by the light intensity power adjustment unit is a square wave pulse width modulation signal, and the duty cycle adjustment range of the pulse width modulation signal is limited to a closed interval of 1% to 99%.

[0038] Example 1:

[0039] The mutation threshold constant of 15 lx / ms and the steady-state counting threshold of 500 cycles were determined based on actual measurement data from the workshop. Under normal unobstructed operation, the slow light intensity drift gradient caused by environmental noise and temperature drift does not exceed 8 lx / ms. When the material conveying unit shuttles at a speed of 1.5 m / s, the illuminance change gradient caused by physical shading exceeds 25 lx / ms. Therefore, the mutation threshold constant is set to 15 lx / ms, placing it between the upper limit of normal fluctuation (8 lx / ms) and the lower limit of shading mutation (25 lx / ms). When the mutation threshold constant is below 8 lx / ms... Regular small adjustments can easily trigger the second control flow, causing the proportional-integral closed loop to disconnect prematurely. When the mutation threshold constant is higher than 25 lx / ms, it cannot identify occlusion interference in time. When the sampling period is 10 ms, 500 consecutive operating cycles correspond to 5 seconds, which can cover the thermal equilibrium transition stage after the light source is continuously powered on. When the steady-state count threshold is lower than 200 cycles, the heat accumulation is not obvious, which can easily cause frequent and ineffective gain corrections. When the steady-state count threshold is higher than 1000 cycles, it cannot correct the steady-state illuminance deviation caused by temperature accumulation in time.

[0040] When the characterization state is the first value 0, the program control unit activates the first control flow, calculates the real-time illuminance residual between the steady-state illuminance data of the current sampling period and the set target value based on the standard proportional-integral gain coefficient, and outputs a control signal to the light intensity power adjustment unit. This control signal is a square wave pulse width modulation signal, and its duty cycle is limited to a closed range of 1% to 99% to maintain the steady-state illuminance of the exposure working surface. When the material conveying unit quickly passes through the exposure area and the characterization state switches to the second value 1, the program control unit activates the second control flow, disconnects the proportional-integral calculation loop and suspends the feedback operator within the transient time window. The integral accumulation caused by the step change in real-time illuminance residual is stopped. Simultaneously, the control signals output to the light intensity power adjustment unit during the 10 consecutive steady-state operating cycles before the state switch are read from the buffer. The arithmetic mean of these control signals is calculated and used as a static constant. The program control unit rewrites the register value of the light intensity power adjustment unit using this static constant, ensuring that the light source maintains its original power output during the transition phase. The width of the transient time window is 3 sampling periods, or 30ms.

[0041] When the number of consecutive operating cycles with the characterization state of 0 reaches 500 cycles, the program control unit obtains the temperature sensor reading from the temperature sensor of the current environment, smoothly converts it into the temperature drift parameter of the current environment, and then makes a slight correction to the standard proportional-integral gain coefficient based on the temperature drift parameter, and limits the correction range to within ±2% of the initial value of the standard proportional-integral gain coefficient, so as to reduce the temperature drift deviation caused by the continuous operation of the light-emitting device.

[0042] When the characterization state switches to the second value 1 and the proportional-integral calculation loop is disconnected, the independent safety detection module acquires the absolute illuminance value from the illuminance detection channel within a 30ms transient time window, according to a 10ms control cycle, and continuously monitors it. The safety detection module calculates the absolute value of the difference between the absolute illuminance value and the set target value, and then divides this absolute value by the set target value to obtain the deviation ratio of the absolute illuminance value relative to the set target value. When the deviation ratio exceeds the safety boundary threshold... At this time, the program control unit forcibly interrupts the signal output of the second control flow, resetting the control signal sent to the light intensity power adjustment unit to a 50% fixed duty cycle safety signal, so that the light source output power is kept within a safe range. The safety boundary threshold is set to 5% of the target value.

[0043] After the material conveying unit leaves the exposure area and the pulsed external interference lasting 30ms disappears, the absolute value G of the time-varying illuminance gradient falls back to below 15lx / ms, the characterization state is reset to the first value 0, and the program control unit re-activates the first control flow. This process uses the response delay of the light intensity power adjustment unit itself to weaken the transient light intensity impact, and switches between continuous feedback control and discrete feedforward backoff according to the time-varying illuminance gradient, so that the system overshoot in the transition phase is reduced from 12% to less than 0.3%, the response delay of the control loop is shortened to 30ms, and finally the stable output of the square wave pulse width modulation signal is restored.

[0044] Example 2: In a specific implementation scenario, the material conveying unit physically interferes with multiple light sources during its passage, causing the working surface of the exposure area to be subjected to transient step stray light disturbances of up to 35 lx / ms. Since the program segment step switching and continuous error feedback in the conventional control loop are on the same time scale, the expected changes in electrical parameters caused by the program step overlap with the external transient strong interference and cannot be independently identified. The proportional-integral operator in the feedback loop continuously accumulates the mutation amount based on the instantaneously decreasing sampled data and outputs adjustment commands with large fluctuations. This causes the light intensity power adjustment unit to experience response lag and alternating overshoot oscillations after the material conveying unit passes through. The transient illuminance fluctuation in the exposure area exceeds the ±0.5% range allowed by the process, thereby reducing the yield of the exposure process.

[0045] To suppress steady-state control failure caused by impulse disturbances, the program control unit uses a fixed sampling period of 10ms. The system continuously acquires illuminance sampling sequences from the illuminance detection channel and feeds them into a moving average filtering time window with a length of 5 sampling points. The program control unit calculates the arithmetic mean of five consecutive illuminance samples within the window, including the current sampling period. This arithmetic mean is used as the steady-state illuminance data for the current sampling period. Then, the steady-state illuminance data for the current sampling period is first-order differencing the steady-state illuminance data from the previous sampling period. After conversion according to a fixed sampling period, the absolute value is taken to obtain the absolute value G of the time-varying illuminance gradient. Finally, G is compared with a constant abrupt change threshold of 15 lx / ms. Compare when G is less than or equal to When G is greater than 0, the representation state is established as the first value 0; when G is greater than 0, the representation state is established as the first value 0. At that time, the representation state is established as the second value 1.

[0046] When the characterization state is the first value 0, the program control unit activates the first control flow, calculates the real-time illuminance residual between the steady-state illuminance data of the current sampling period and the set target value based on the standard proportional-integral gain coefficient, and outputs a control signal to the light intensity power adjustment unit. This control signal uses a square wave pulse width modulation signal with a duty cycle limited to a closed interval of 1% to 99%. When the characterization state switches to the second value 1, the program control unit activates the second control flow, disconnects the proportional-integral calculation loop and suspends the feedback operator within a 30ms transient time window. The integral accumulation caused by the step change in real-time illuminance residual is stopped. At the same time, the control signals output to the light intensity power adjustment unit in the 10 consecutive steady-state operation cycles before the state switch are read from the buffer. The arithmetic mean of these control signals is calculated and written as a static constant into the register value of the light intensity power adjustment unit, so that the light source maintains the original power output during the transition phase.

[0047] To verify the impact of different control paths on the illuminance adjustment process, under normal low-disturbance conditions, the disturbance change rate of the experimental group was 2.1 lx / ms, and the absolute value G of the time-varying illuminance gradient calculated by the program control unit was 1.8 lx / ms, which is lower than the abrupt change threshold constant. Therefore, the characterization state is set to the first value of 0, and the proportional-integral closed loop is kept active. Under this condition, the maximum overshoot in the transition section is 0.05%, the response delay is 10ms, and the steady-state convergence residual is 0.12%. Under the same disturbance change rate of 2.1lx / ms, the absolute value of the time-varying illuminance gradient calculated by the control group is 1.9lx / ms, but no characterization state shunting is set, and the continuous proportional-integral closed loop is still used. Its maximum overshoot in the transition section is 0.06%, the response delay is 10ms, and the steady-state convergence residual is 0.12%.

[0048] Under transient strong disturbance conditions, the disturbance change rate of the experimental group increased to 35.0 lx / ms, and the calculated absolute value of the time-varying illuminance gradient G was 33.4 lx / ms. Since this value is greater than the abrupt change threshold constant of 15 lx / ms, The program control unit switches the characterization state to the second value 1, disconnects the proportional-integral calculation loop, and writes the average value of the historical control signal as a static constant into the register value of the light intensity power adjustment unit. After this processing, the maximum overshoot of the transition section is 0.28%, the response delay is 30ms, and the steady-state convergence residual is 0.14%.

[0049] The control group calculated an absolute value of the time-varying illuminance gradient of 33.8 lx / ms under a perturbation rate of 35.0 lx / ms. However, since no state-characterizing shunt was set, the continuous proportional-integral closed loop continued to run. After the transiently decreasing illuminance sample value continued to participate in the calculation of the real-time illuminance residual, the maximum overshoot in the transition section reached 12.40%, the response delay was extended to 145 ms, and the steady-state convergence residual was 0.48%. This set of data shows that under similar absolute values ​​of the time-varying illuminance gradient, the continuous proportional-integral closed loop will continue to accumulate the real-time illuminance residual during the occlusion period and form a large transition section output after the occlusion disappears.

[0050] The second control group was also subjected to a disturbance rate of 35.0 lx / ms, and the calculated absolute value of the time-varying gradient of illuminance was 33.1 lx / ms. The characterization state was switched to the second value 1. After disconnecting the proportional-integral calculation loop, this group output a zero duty cycle signal, but did not use the average of historical control signals to maintain the light source power. The maximum overshoot in the transition section was 8.65%, the response delay was 90 ms, and the steady-state convergence residual was 0.35%. It can be seen that directly outputting a zero duty cycle in the second control flow will cause the light source power to decrease within the transient time window, and a stable output still needs to be re-established when the closed-loop regulation is restored.

[0051] To verify the effect of the mutation threshold constant on the switching of the characterization state, control group 3 was subjected to a perturbation rate of 6.8 lx / ms, and the calculated absolute value of the time-varying illuminance gradient was 6.2 lx / ms. This value did not reach 15 lx / ms, but this group incorrectly switched the characterization state to the second value 1, erroneously triggering the feedback blocking management. The maximum overshoot of the resulting transition segment was 3.15%, the response delay was 80 ms, and the steady-state convergence residual was 1.85%. Control group 4 was subjected to a perturbation rate of 35.0 lx / ms, and the calculated absolute value of the time-varying illuminance gradient was 33.6 lx / ms, but the characterization state was still kept at the first value 0, and the second control flow was not triggered. The proportional-integral calculation loop continued to adjust according to the real-time illuminance residual after the mutation. The maximum overshoot of the resulting transition segment was 11.80%, the response delay was 140 ms, and the steady-state convergence residual was 0.52%.

[0052] When the number of consecutive operating cycles with the characterization state at the first value of 0 reaches 500 cycles, the program control unit obtains the temperature sensor reading from the temperature sensor of the current environment, smoothly converts it into temperature drift parameters, and makes a slight correction to the standard proportional-integral gain coefficient based on these parameters. The correction range is limited to ±2% of the initial value of the standard proportional-integral gain coefficient. After the test group has run for 500 consecutive cycles, the applied disturbance change rate is 1.2 lx / ms, the calculated absolute value of the time-varying illuminance gradient is 1.1 lx / ms, the characterization state remains at the first value of 0, and the slight correction of the standard proportional-integral gain coefficient is triggered. Under this condition, the maximum overshoot in the transition section is 0.04%, the response delay is 10 ms, and the steady-state convergence residual is 0.08%.

[0053] When the characterization state is the second value 1 and the proportional-integral calculation loop is in the open state, the independent safety detection module acquires the absolute illuminance value from the illuminance detection channel within a 30ms transient time window and continuously monitors it according to a 10ms control cycle. The safety detection module calculates the deviation ratio of the absolute illuminance value from the set target value; when this deviation ratio exceeds the 5% safety boundary threshold... At this time, the program control unit forcibly interrupts the signal output of the second control flow and resets the control signal sent to the light intensity power adjustment unit to a 50% fixed duty cycle safety signal, so that the output power of the light source is kept within a safe range.

[0054] After the material conveying unit leaves the exposure area and the pulsed external interference lasting 30ms disappears, the absolute value G of the time-varying illuminance gradient falls back to below 15lx / ms, the characterization state is reset to the first value 0, and the program control unit reactivates the first control flow. This control process uses the response delay of the light intensity power adjustment unit itself to weaken the transient light intensity impact, and switches between continuous feedback control and discrete feedforward yield according to the time-varying illuminance gradient, so that the system overshoot in the transition phase is reduced from 12% to less than 0.3%, the response delay of the control loop is shortened to 30ms, and finally the stable output of the square wave pulse width modulation signal is restored.

[0055] Example 3: The current program control unit is connected to the register port of the optical intensity power adjustment unit through the pulse width modulation output channel. The register value of the optical intensity power adjustment unit is carried by the compare-capture register of the general-purpose timer. The master control state machine inside the program control unit runs according to an increasing time sequence and has a discrete control sampling period of 10ms. A hardware interrupt is triggered periodically to read the illuminance sampling sequence from the analog-to-digital converter register of the illuminance detection channel. The main control state machine moves the latest 5 illuminance sampling values ​​into the circular buffer, calculates the arithmetic mean of the 5 consecutive illuminance sampling values, filters out electromagnetic interference components in the input channel, and uses the arithmetic mean as the steady-state illuminance data for the current sampling period.

[0056] The master control state machine reads the steady-state illuminance data from the previous sampling period, performs a first-order difference between it and the steady-state illuminance data of the current sampling period, takes the absolute value of the first-order difference, and converts it according to a fixed sampling period to obtain the absolute value G of the time-varying illuminance gradient for the current sampling period. Then, G is compared with a pre-written 15 lx / ms abrupt change threshold constant. Compare when G is less than or equal to When G is greater than 0, the master control state machine sets the state representation in the state register to the first value 0; when G is greater than 0, the master control state machine sets the state representation in the state register to the first value 0. At that time, the characterization state is set to the second value 1, the counter is started to accumulate the number of discrete cycles, and the transient time window is set. Limited to 30ms.

[0057] During the transient time window when the material conveying unit blocks the light path, the decrease in illuminance on the working surface is due to the mechanical blockage of the physical light path. At this time, the second control flow does not increase the light source power to compensate for the blocked luminous flux. Instead, it disconnects the proportional-integral calculation loop and suspends the feedback operator, locking the state of the error accumulator inside the controller. At the same time, it uses the average of historical control signals to maintain the light source output. When continuous feedback control continues during the blockage period, the low illuminance data collected by the illuminance detection channel will increase the real-time illuminance residual, causing the proportional term to rise and the integral term to saturate. After the material conveying unit leaves the exposure area within 30ms, the accumulated integral term will cause the control loop to output an excessively high drive duty cycle, resulting in overshoot oscillation in the transition section. During the 30ms transient time window, the second control flow maintains the control signal at the average output level of the 10 consecutive steady-state operating cycles before the blockage, keeping the luminous flux of the light source itself stable. When the blockage disappears, the illuminance on the working surface recovers to near the initial steady-state level. The secondary overshoot caused by integral saturation is suppressed, and the system overshoot is controlled within 0.3%.

[0058] When the status register is 0, the program control unit activates the first control flow and runs the proportional-integral calculation loop. It subtracts the set target value from the steady-state illuminance data of the current sampling period to obtain the real-time illuminance residual. The program control unit multiplies the real-time illuminance residual by the proportional coefficient and the integral coefficient to obtain the proportional term and the integral term. After merging the proportional term and the integral term, a square wave pulse width modulation signal is generated and output to the light intensity power adjustment unit. The duty cycle of the square wave pulse width modulation signal is limited to a closed range of 1% to 99%. Its duty cycle value is written into the general-purpose timer comparison capture register used as the light intensity power adjustment unit register to adjust the illuminance of the exposure area working surface. At the same time, the program control unit writes the duty cycle value of the current control signal into the dynamic update queue. The dynamic update queue adopts a first-in-first-out buffer queue with a depth of 10 to always store the control signal for 10 consecutive steady-state operating cycles.

[0059] When the representation state switches to the second value 1, the program control unit activates the second control flow, stops the update of the accumulator in the proportional-integral calculation loop and locks the current integral term value, thus preventing integral saturation caused by the step change in real-time illuminance residual, within a 30ms transient time window. Inside, the program control unit reads all control signals from the dynamic update queue for 10 consecutive steady-state operating cycles, calculates their arithmetic mean, and determines the arithmetic mean as a static constant. Subsequently, the static constant is directly overwritten into the general-purpose timer comparison capture register, which serves as the light intensity power adjustment unit register, through the internal data bus, so that the light source luminous power remains stable during the transition phase.

[0060] During the period in which the first control flow remains on and the characterization state remains at the first value of 0, the counter accumulates the number of steady-state operating cycles cycle by cycle. When the counter reaches the steady-state counting threshold of 500 cycles, the program control unit reads the input signal voltage value from the temperature sensor channel set on the light source base, and performs a smooth conversion on the input signal voltage value according to the calibrated linear voltage-temperature conversion coefficient to obtain the temperature drift parameter of the current environment. The program control unit calculates the adjustment component based on the temperature drift parameter and performs a slight correction on the standard proportional-integral gain coefficient in the first control flow. The correction range of the proportional coefficient and the integral coefficient is limited to within ±2% of their respective initial values ​​to correct the efficiency drift caused by the continuous heating of the light-emitting device.

[0061] During the 30ms transient time window when the characterization state is the second value 1 and the proportional-integral calculation loop is in a suspended state... Inside, an independent safety detection module continuously monitors the absolute illuminance value output by the illuminance detection channel according to a 10ms control cycle. The safety detection module calculates the absolute value of the difference between the absolute illuminance value and the set target value, and then compares this absolute value with the set target value to obtain the deviation ratio of the absolute illuminance value relative to the set target value. When this deviation ratio exceeds the 5% safety boundary threshold... At this time, the safety detection module generates a hardware unmasked interrupt signal, seizes the bus control of the main control chip, forcibly cuts off the overwriting action of the second control flow on the light intensity power adjustment unit register, resets the channel output state of the general timer, and sends a 50% fixed duty cycle safety signal to the light intensity power adjustment unit to limit the overload of the light-emitting element.

[0062] When the light source hardware is damaged, the illuminance detection channel malfunctions, or the occlusion time exceeds the predetermined 30ms transient time window, the deviation of the absolute illuminance value from the set target value may continue to exceed 5%. Once this deviation exceeds the safety boundary threshold, the independent safety detection module takes over control through a hardware non-shielded interrupt. During the occlusion process of the material conveying unit's normal shuttle, the deviation of the absolute illuminance value remains within the 5% safety boundary threshold. The second control flow continues to use the historical control signal average to maintain the light source output. The 50% fixed duty cycle safety signal corresponds to the rated central safety power point of the light-emitting device and is used to prevent the drive circuit from experiencing thermal overload during the proportional-integral calculation loop suspension. When this safety signal is issued, the pulse width modulation module inside the main control chip enables the single-edge slope limiting logic to smoothly transition the duty cycle to 50% within 1ms, reducing the step impact caused by the sudden change in duty cycle.

[0063] When the interfering shadow moves out of the exposure area, the data collected by the illuminance detection channel, after first-order difference, absolute value taking, and sampling period conversion, results in the illuminance time-varying gradient absolute value G dropping below 15 lx / ms. The main control state machine clears the counter, resetting the characterization state to the first value 0. The program control unit restores the update state of the proportional-integral calculation loop, reconnects the first control flow, and continues to perform closed-loop illuminance adjustment. After the switch is completed, the amplitude of the illuminance pulse fluctuation in the transition section is limited to within 0.3%, the system's action response delay is maintained at 30 ms, and the illuminance on the working surface is restored to the error range corresponding to the set target value.

[0064] Example 4: This example combines Figures 1 to 2 This paper describes a programmable control method for illuminance regulation, such as... Figure 1 As shown, the control process sequentially includes step S1 of acquiring steady-state data of the sequence filter output, step S2 of calculating the first-order difference to determine the absolute value of the gradient, step S3 of comparing the gradient with the threshold to determine the split state, and step S4 of allocating the path output signal according to the characterization state.

[0065] like Figure 2 As shown, the first control flow state executes closed-loop output and maintains the queue, and maintains its own state when the steady-state counting gain is reached. When the absolute value of the gradient is greater than the mutation threshold, the first control flow state switches to the second control flow state. At this time, the feedback loop is disconnected, and the system is managed by the average of historical control signals for output. When the absolute value of the second control flow state is less than or equal to the threshold, it switches back to the first control flow state. When the deviation of the second control flow state exceeds the boundary threshold, it switches to the corresponding forced interrupt safety state that resets the fixed duty cycle output. When the forced interrupt safety state recovers to the steady state after the gradient falls back, it switches back to the first control flow state.

[0066] Example 5: Before the program control unit is connected to the illuminance detection channel in the exposure workshop, the calibration-grade illuminance meter and photoelectric sensor are placed in an unobstructed standard light field. The light intensity power adjustment unit is driven to increase the luminous power step by step in the duty cycle range of 1% to 99%. At the same time, the voltage value output by the analog-to-digital converter register of the illuminance detection channel is recorded. Based on the absolute illuminance value measured by the calibration-grade illuminance meter, a linear calibration ratio coefficient between the voltage value and the absolute illuminance value is established.

[0067] During the initialization phase of the program control unit, five initial illuminance sample values ​​are continuously collected according to a discrete control sampling period of 10ms. These values ​​are written into a circular buffer, and the arithmetic mean is calculated as the initial steady-state illuminance data. The program control unit then writes the corresponding initial square wave pulse width modulation signal duty cycle value ten times to fill the dynamic update queue using a first-in-first-out method. This ensures that the initial calculation cycle has complete historical control signals, avoids misjudgment of the absolute value of the time-varying illuminance gradient due to missing data, and establishes the representation state in the status register as the first value of 0.

[0068] Before system deployment, a temperature gradient of 25°C to 45°C is applied to the control cabin. The light intensity attenuation slope of the light-emitting device at different temperatures is measured, and the correction range of the standard proportional-integral gain coefficient is limited to within ±2% of the initial value to reduce the steady-state illuminance residual caused by excessive correction of the temperature drift parameter. The program control unit subtracts the reference calibration temperature from the real-time acquired temperature drift parameter to obtain the absolute value of the temperature deviation. Then, the absolute value of the temperature deviation is multiplied by the pre-calibrated light intensity temperature attenuation coefficient to obtain the proportional-integral gain correction amount. Subsequently, the proportional-integral gain correction amount is accumulated to the standard proportional gain coefficient and the standard integral gain coefficient in the same proportion. When the corrected gain coefficient exceeds the +2% or -2% boundary of its initial value, it is truncated to the corresponding boundary value to complete the compensation for the thermal drift of the light-emitting device.

[0069] Before the system starts closed-loop regulation, the main control state machine sends a simulated pulse over-limit signal to the independent safety detection module to verify the preemption response delay of the hardware unmasked interrupt to the second control flow. When the deviation ratio of the absolute illuminance value corresponding to the simulated sample value from the set target value reaches the safety boundary threshold of 5%, the independent safety detection module resets the pulse width modulation output channel in less than 1ms and outputs a 50% fixed duty cycle safety signal to the light intensity power adjustment unit, thereby completing the full-link safety boundary verification of the control loop.

[0070] Example 6: The general-purpose timer comparison capture register, which serves as the register of the light intensity power adjustment unit, has a microsecond-level response delay during the continuous periodic high-frequency level switching process. As heat continues to accumulate inside the driving component, this response delay changes nonlinearly, causing a deviation between the control parameters written by the main control state machine and the actual output of the light intensity power adjustment unit. During the discrete adjustment process of the semiconductor exposure process, the material conveying unit shuttles frequently according to a fixed step, and the baseline of the ambient background illuminance changes discretely. During the first control flow connection period, the program control unit reads the current actual output waveform feedback through the capture channel of the general-purpose timer.

[0071] The master control state machine calculates the target high-level duration based on the target set duty cycle and the total duration of a single cycle of the square wave pulse width modulation signal written into the general-purpose timer comparison capture register. Then, it subtracts the target high-level duration from the actual high-level duration captured to obtain the time delay deviation increment. Its positive or negative sign is used to determine the direction of time delay compensation. The program control unit will... Write a delay integrator counter to record the delay changes of the general-purpose timer output waveform within a continuous control cycle, and correct the actual output deviation formed by the first control flow after the standard proportional-integral gain coefficient is calculated.

[0072] The program control unit will measure the time delay deviation increment. Divide by the total duration of a single cycle of the current square wave pulse width modulation signal to obtain a dimensionless time delay ratio value. Then multiply this time delay ratio value by a preset reference duty cycle conversion factor to obtain a duty cycle compensation correction operator. During the first control flow operation, the master control state machine corrects the triggering time of the general-purpose timer output edge according to the duty cycle compensation correction operator, so that the actual output waveform corresponds to the target set duty cycle written into the general-purpose timer compare capture register.

[0073] When the absolute value of the time-varying gradient of illuminance, G, crosses the abrupt change threshold constant of 15 lx / ms When the representation state switches to the second value 1, the program control unit activates the second control flow, disconnects the proportional-integral calculation loop and suspends the feedback operator. The program control unit reads the control signal from the dynamic update queue, calculates the average value of the historical control signal and uses it as a static constant. This static constant is then directly written into the register value of the light intensity power adjustment unit via the internal data bus within a 30ms transient time window. Inside, the write value of the register remains the static constant. The duty cycle compensation correction operator is only used to correct the actual triggering time of the general-purpose timer output edge to compensate for the microsecond-level level delay caused by high-frequency switching and to maintain the luminous power of the light source during transient hosting.

[0074] After the material conveying unit leaves the exposure area and the 30ms transient time window ends, the absolute value G of the time-varying illuminance gradient falls back to the abrupt change threshold constant. Within this range, the main control state machine resets the characterization state to the first value 0, clears the delay integrator counter, and returns control to the first control flow. The program control unit restores the update state of the proportional-integral calculation loop. During the transition phase, the overshoot of the working surface illuminance is kept within 0.3%, and the duty cycle of the square wave pulse width modulation signal of the general-purpose timer output channel converges to the target set value.

[0075] 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 programmable control method for illuminance regulation, characterized in that, Includes the following steps: Step S1: The program control unit acquires the illuminance sampling sequence from the illuminance detection channel at a fixed sampling period, smooths the illuminance sampling sequence using a moving average filtering algorithm, and outputs the steady-state illuminance data for the current sampling period. Step S2: The program control unit calculates the first-order difference between the steady-state illuminance data of the current sampling period and the steady-state illuminance data of the adjacent previous sampling period to determine the absolute value of the time-varying gradient of illuminance. Step S3: The program control unit compares the absolute value of the time-varying illuminance gradient with a preset mutation threshold constant to determine the control shunting state: when the absolute value of the time-varying illuminance gradient is less than or equal to the mutation threshold constant, the characterization state is established as the first value; when the absolute value of the time-varying illuminance gradient is greater than the mutation threshold constant, the characterization state is established as the second value. Step S4: The program control unit allocates a control path based on the characterization state: when the characterization state is the first value, the first control flow is activated, and the real-time illuminance residual between the steady-state illuminance data of the current sampling period and the set target value is calculated based on the standard proportional-integral gain coefficient, and the control signal is output to the light intensity power adjustment unit; when the characterization state is the second value, the second control flow is activated, the proportional-integral calculation loop is disconnected and the feedback operator is suspended, and the average value of the historical control signal is extracted as a static constant within the set transient time window, and the register value of the light intensity power adjustment unit is rewritten through the static constant.

2. The program control method for illuminance regulation according to claim 1, characterized in that, The sampling period is 10ms; in the moving average filtering algorithm, the program control unit extracts the illuminance sample values ​​of 5 consecutive sampling periods including the current sampling period, calculates the arithmetic mean of the illuminance sample values ​​of 5 sampling periods, and establishes the arithmetic mean as the steady-state illuminance data of the current sampling period; the mutation threshold constant is 15lx / ms; the width of the transient time window is 3 sampling periods, corresponding to a time length of 30ms; the average of the historical control signals is the average of the control signals output to the light intensity power adjustment unit in the 10 consecutive steady-state operation periods before the current sampling period.

3. The program control method for illuminance regulation according to claim 1, characterized in that, When the number of consecutive operating cycles with the first value representing the state reaches the set steady-state counting threshold, the path for enabling the first control flow in step S4 also includes the following sub-steps: Step S41, the program control unit obtains the temperature sensor reading from the temperature sensor of the current environment and smoothly converts it into the temperature drift parameter of the current environment; Step S42, the program control unit makes a slight correction to the standard proportional-integral gain coefficient based on the temperature drift parameter, limiting the correction range to within ±2% of the initial value of the standard proportional-integral gain coefficient, so that the steady-state illuminance data is stabilized within the error range corresponding to the set target value.

4. The program control method for illuminance regulation according to claim 1, characterized in that, Step S4 includes the following sub-steps during the period when the proportional-integral calculation loop of the second control flow is disconnected: Step S43, the independent safety detection module obtains the absolute illuminance value from the illuminance detection channel at a control cycle of 10ms and continuously monitors the absolute illuminance value to determine whether the absolute illuminance value deviates from the set target value; Step S44, when the absolute illuminance value deviates from the set target value by more than the set safety boundary threshold, the program control unit forcibly interrupts the signal output of the second control flow and resets the control signal sent to the light intensity power adjustment unit to the set 50% fixed duty cycle safety signal.

5. The program control method for illuminance regulation according to claim 1, characterized in that, The path for enabling the first control flow in step S4 includes the following sub-steps: Step S45, the program control unit subtracts the set target value from the steady-state illuminance data of the current sampling period to obtain the current real-time illuminance residual; Step S46, the program control unit multiplies the real-time illuminance residual by the proportional coefficient and the integral coefficient respectively to obtain the proportional term and the integral term, and combines the proportional term and the integral term to generate a control signal for driving the light intensity power adjustment unit.

6. The program control method for illuminance regulation according to claim 1, characterized in that, The path for extracting the historical control signal mean in step S4 includes the following sub-steps: Step S47, when the representation state is the first value, the program control unit continuously writes the control signal output in each sampling period into the buffer, maintaining a dynamic update queue of control signals containing 10 consecutive steady-state cycles; Step S48, when the representation state switches to the second value, the program control unit reads all control signals in the dynamic update queue, calculates the arithmetic mean of all control signals, and establishes the arithmetic mean as a static constant.

7. The program control method for illuminance regulation according to claim 3, characterized in that, The steady-state counting threshold is 500 consecutive operating cycles.

8. The program control method for illuminance regulation according to claim 4, characterized in that, The safety boundary threshold is 5% of the set target value.

9. The program control method for illuminance regulation according to claim 1, characterized in that, The control signal received by the optical intensity power adjustment unit is a square wave pulse width modulation signal, and the duty cycle adjustment range of the pulse width modulation signal is limited to a closed interval of 1% to 99%.

Images