Instruction distribution logic control system for combustion engine monitoring environment

Abstract

The application relates to the field of industrial control and discloses an instruction distribution logic control system used in a gas turbine monitoring environment, which comprises an instruction encapsulation module, a redundant transmission module and a terminal control module. The instruction encapsulation module is used for determining a dynamic response characteristic vector according to a gas turbine physical parameter and encapsulating the dynamic response characteristic vector with an operation instruction to form a monitoring message. The redundant transmission module is used for forwarding the monitoring message in a switching period. The terminal control module is used for completing kinematic compensation prediction by using the dynamic response characteristic vector when a message time difference exceeds a threshold value, driving a controlled terminal to act, and controlling a predicted track to complete damping convergence by using a transition correction coefficient after the message is recovered. The application integrates physical inertia into the message, converts passive reception into autonomous prediction mode during flow interruption of the controller, eliminates the coupling between network jitter and mechanism action, guarantees smooth operation of equipment, and prevents mechanical impact.

Description

Technical Field

[0001] This invention relates to a command distribution logic control system for gas turbine monitoring environments, belonging to the field of industrial control technology. Background Technology

[0002] Currently, in power system dispatch and monitoring, remote distribution of control commands via optical transmission networks is fundamental to ensuring the coordinated operation of power plant equipment. Existing Automatically Switched Optical Networks (ASON) ensure communication link connectivity through path protection and rerouting mechanisms. This technology achieves millisecond-level service recovery in scenarios of fiber optic interruption or node failure through dynamic resource scheduling at the optical layer. However, gas turbine monitoring processes have high isochronous control requirements. The gas turbine servo control cycle is typically set to 10ms, requiring control commands to have a definite arrival sequence. In a wide area network transmission environment, even if ASON technology compresses the rerouting switching cycle to within 50ms, several control cycle command gaps still occur at the physical execution level. Since the communication transmission layer and the physical control layer are logically decoupled, the timing jitter caused by link rerouting can distort the arrival pattern of commands on the terminal side.

[0003] To address these challenges, the industry has attempted to shorten link switching time through hardware upgrades. Analysis reveals that the convergence speed of optical layer cross-connections and routing protocols is limited by the physical hardware response limits, making it difficult to completely eliminate latency blind spots exceeding microseconds. Introducing a data buffering mechanism at the terminal side increases the inherent closed-loop latency of the control loop, disrupting the real-time performance of the gas turbine regulation process and even inducing step oscillations in the servo valves. Existing command distribution architectures cannot maintain the continuity of physical actuator actions in dynamic network fluctuation environments. Besides hardware limitations, control method logic strategies also have shortcomings in maintaining the continuity of mechanism movement. For example, Chinese invention patent application CN121603359A... A link switching method and system for complex multi-machine, multi-plane networks are disclosed. The method optimizes node resource allocation through dynamic programming algorithm and completes automatic link switching based on priority order and load status. The essence lies in the optimization of network transmission layer connectivity and load balancing. In the highly dynamic physical scenario of gas turbine control, the existing technology only focuses on the rapid recovery of data channels and does not establish the coupling relationship between communication status and the physical inertia of the controlled terminal. Due to the lack of real-time perception of mechanical motion law, during the vacuum period of link rerouting control, the terminal controller loses feedforward excitation and enters a runaway or maintenance state. It cannot resist the inertial offset within the mechanical time constant. After communication is restored, the instantaneous jump of command value induces step oscillation of servo valve or unplanned tripping.

[0004] Therefore, how to solve the mechanical impact on the gas turbine actuator caused by the command interruption resulting from wide area network link switching while maintaining communication connectivity has become the technical problem to be solved by this invention. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: A command distribution logic control system for a gas turbine monitoring environment, the system comprising: The instruction encapsulation module is used to obtain the gas turbine rotor speed, load power, and historical displacement parameters of the execution terminal, and to determine the dynamic response feature vector based on the angular acceleration of the rotor speed, the instantaneous rate of change of the load power, and the motion gradient of the historical displacement parameters. The instruction encapsulation module binds the dynamic response feature vector as a physical inertia constraint to the operation instruction in a time sequence and encapsulates it into a monitoring message with source physical attribute identifier. The redundant transmission module, connected to the instruction encapsulation module, is used to build a primary and backup dual-plane communication link based on the automatic switching optical network. Within a 50ms switching period after the primary link failure is detected, the monitoring message is redirected to the backup link for distribution through the routing protection algorithm. The end-control module, connected to the redundant transmission module, is used to monitor the time difference between the reception of two adjacent monitoring messages. ; in the receiving time difference When the preset message reception time threshold is exceeded, the terminal control module uses the dynamic response feature vector in the monitoring message to start the kinematic compensation prediction process, calculates and generates a predictive control trajectory that satisfies the mechanical time constant of the gas turbine, and drives the execution terminal to complete smooth displacement; after the monitoring message resumes normal reception, it obtains the control quantity deviation between the predictive control trajectory and the real-time operation command, and uses the transition correction coefficient dynamically updated according to the control quantity deviation to control the predictive control trajectory to complete damped convergence to the operation command.

[0006] Preferably, the end control module is equipped with a prediction compensation unit; the prediction compensation unit extracts the rotor speed fluctuation envelope and load disturbance amplitude from the dynamic response feature vector in real time, and completes the second derivative fitting in combination with the operation command in the last valid message to determine the displacement increment compensation value of the execution terminal during the communication switching vacuum period, so as to maintain the second-order continuity of the controlled physical state quantity.

[0007] Preferably, when the redundant transmission module initiates the link protection switching logic, it synchronously sends a path switching status message to the end control module; after receiving the path switching status message, the end control module reduces the message reception time threshold to improve the end control module's sensitivity to wide area network transmission jitter and packet loss events.

[0008] Preferably, the dynamic response feature vector includes the equivalent inertia constant of the execution terminal; the end control module determines the upper limit of the rate of change of the predicted control trajectory based on the equivalent inertia constant, ensuring that the increment of the generated virtual control quantity does not exceed the physical response envelope of the execution terminal, and avoiding nonlinear oscillations in the gas turbine combustion chamber induced by command jumps.

[0009] Preferably, if the absolute value of the displacement deviation between the predicted control trajectory and the real-time issued operation command increases, the end control module increases the value of the transition correction coefficient to extend the logical approximation period of the execution terminal from the local autonomous prediction mode to the remote real-time control mode, thereby offsetting the step impact of the command value caused by the link switching.

[0010] Preferably, the redundant transmission module determines the continuity of the message by verifying the cyclic sequence count value in the header of the monitoring message protocol; if the cumulative transmission duration corresponding to the number of interruptions in the cyclic sequence count value exceeds the message reception time threshold, the end control module automatically switches to the autonomous logic processing state based on the dynamic response feature vector.

[0011] Preferably, the dynamic response feature vector is associated with a load power envelope that characterizes the thermodynamic features of the gas turbine; the end control module performs dynamic gain correction on the predicted control trajectory based on the load power envelope to offset the rotor speed frequency drift caused by the transient load on the grid side, and to ensure the physical self-consistency of the gas turbine control logic.

[0012] Preferably, the end control module calculates the phase offset vector between the predicted control trajectory and the restored operation command in real time, and adjusts the message reception time threshold of the next statistical period according to the phase offset vector to adaptively match the asymmetric transmission delay differences caused by different physical path switching.

[0013] Preferably, after completing the communication path switching action, the redundant transmission module sends a register reset signal to the end control module; the end control module responds to the register reset signal, clears the prediction cache stack established based on the dynamic response feature vector, and re-completes the synchronization reference calibration of the system logic clock.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. In the instruction distribution logic control, this scheme encapsulates the execution inertia envelope vector representing the gas turbine rotor speed, load power, and servo valve stroke in the distributed control instruction frame. This enables the end controller to switch from a simple passive receiving mode to an autonomous reasoning mode based on the underlying physical kinematics during instruction timing discontinuities caused by wide-area communication link switching. This eliminates the direct coupling between network transmission layer latency jitter and physical actuator actions at the control logic level, ensuring the operational stability of high-speed rotating equipment during dynamic changes in the communication environment.

[0015] 2. By utilizing the logical judgment between the time difference of adjacent control command frames and the preset isochronism tolerance threshold, and in conjunction with the local logic interpolation mechanism, a virtual simulation control trajectory conforming to the mechanical time constant is generated. This enables the servo actuator to maintain the smooth continuation of physical actions along the simulation trajectory during the interruption of the command flow, avoiding the actuator step oscillation or unplanned tripping caused by the loss of commands during the network switching cycle in the existing technology, and improving the tolerance of the control system to the transient uncertainty of the wide area network.

[0016] 3. By introducing convergence adjustment based on damping smoothness coefficient during the command flow recovery stabilization phase, the controller gradually approximates the actual command displacement after network recovery based on the absolute deviation between the locally extrapolated displacement and the actual command displacement. This eliminates the secondary mechanical shock caused by sudden changes in command values ​​after the main / backup path switch is completed, ensuring a smooth transition of the actuator from local predictive logic to remote real-time control logic. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the instruction distribution logic and autonomous prediction compensation process of this invention. Figure 2 This is a schematic diagram of the hardware architecture and physical state feedback of the system of the present invention.

[0018] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] 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.

[0020] A command distribution logic control system for a gas turbine monitoring environment, the system comprising: The instruction encapsulation module is used to obtain the gas turbine rotor speed, load power, and historical displacement parameters of the execution terminal, and to determine the dynamic response feature vector based on the angular acceleration of the rotor speed, the instantaneous rate of change of the load power, and the motion gradient of the historical displacement parameters. The instruction encapsulation module binds the dynamic response feature vector as a physical inertia constraint to the operation instruction in a time sequence and encapsulates it into a monitoring message with source physical attribute identifier. The redundant transmission module, connected to the instruction encapsulation module, is used to build a primary and backup dual-plane communication link based on the automatic switching optical network. Within a 50ms switching period after the primary link failure is detected, the monitoring message is redirected to the backup link for distribution through the routing protection algorithm. The end control module, connected to the redundant transmission module, is used to monitor the time difference between the reception of two adjacent monitoring messages. When the time difference exceeds the preset message reception time threshold, the end control module uses the dynamic response feature vector in the monitoring message to initiate a kinematic compensation prediction process, calculates and generates a predictive control trajectory that satisfies the mechanical time constant of the gas turbine, and drives the execution terminal to complete smooth displacement. After the monitoring message resumes normal reception, the module obtains the control quantity deviation between the predictive control trajectory and the real-time operation command, and uses the transition correction coefficient dynamically updated according to the control quantity deviation to control the predictive control trajectory to complete damped convergence towards the operation command.

[0021] Preferably, the end control module is equipped with a prediction compensation unit; the prediction compensation unit extracts the rotor speed fluctuation envelope and load disturbance amplitude from the dynamic response feature vector in real time, and completes the second derivative fitting in combination with the operation command in the last valid message to determine the displacement increment compensation value of the execution terminal during the communication switching vacuum period, so as to maintain the second-order continuity of the controlled physical state quantity.

[0022] Preferably, when the redundant transmission module initiates the link protection switching logic, it synchronously sends a path switching status message to the end control module; after receiving the path switching status message, the end control module reduces the message reception time threshold to improve the end control module's sensitivity to wide area network transmission jitter and packet loss events.

[0023] Preferably, the dynamic response feature vector includes the equivalent inertia constant of the execution terminal; the end control module determines the upper limit of the rate of change of the predicted control trajectory based on the equivalent inertia constant, ensuring that the increment of the generated virtual control quantity does not exceed the physical response envelope of the execution terminal, and avoiding nonlinear oscillations in the gas turbine combustion chamber induced by command jumps.

[0024] Preferably, if the absolute value of the displacement deviation between the predicted control trajectory and the real-time issued operation command increases, the end control module increases the value of the transition correction coefficient to extend the logical approximation period of the execution terminal from the local autonomous prediction mode to the remote real-time control mode, thereby offsetting the step impact of the command value caused by the link switching.

[0025] Preferably, the redundant transmission module determines the continuity of the message by verifying the cyclic sequence count value in the header of the monitoring message protocol; if the cumulative transmission duration corresponding to the number of interruptions in the cyclic sequence count value exceeds the message reception time threshold, the end control module automatically switches to the autonomous logic processing state based on the dynamic response feature vector.

[0026] Preferably, the dynamic response feature vector is associated with a load power envelope that characterizes the thermodynamic features of the gas turbine; the end control module performs dynamic gain correction on the predicted control trajectory based on the load power envelope to offset the rotor speed frequency drift caused by the transient load on the grid side, and to ensure the physical self-consistency of the gas turbine control logic.

[0027] Preferably, the end control module calculates the phase offset vector between the predicted control trajectory and the restored operation command in real time, and adjusts the message reception time threshold of the next statistical period according to the phase offset vector to adaptively match the asymmetric transmission delay differences caused by different physical path switching.

[0028] Preferably, after completing the communication path switching action, the redundant transmission module sends a register reset signal to the end control module; the end control module responds to the register reset signal, clears the prediction cache stack established based on the dynamic response feature vector, and re-completes the synchronization reference calibration of the system logic clock.

[0029] Example 1: In an industrial control scenario where heavy-duty gas turbines are dispatched across regions using an automatic switched optical network, a physical interruption of the primary communication link triggers a backup link switchover via a routing protection algorithm. This switchover generates a 50ms communication switching cycle. During this cycle, the control command stream experiences a physical interruption. The conventional controller freezes its output increment due to missing underlying commands, inducing a step-like mechanical shock in the controlled actuator upon communication recovery and causing a frequency drift in the rotor speed on the grid side. During normal operation of the communication link, the command encapsulation module continuously acquires historical parameters of the gas turbine rotor speed, load power, and displacement of the execution terminal. Based on the angular acceleration of the rotor speed, the instantaneous rate of change of the load power, and the motion gradient of the historical displacement parameters, the module determines a dynamic response feature vector. This dynamic response feature vector is then used as a physical inertia constraint and synchronously bound to the operation command sequence, encapsulating it into a monitoring message with source-end physical attribute identification. This dynamic response feature vector is not a simple numerical feedback of local basic physical parameters, but rather a global electromechanical coupling boundary constraint calculated by the command encapsulation module in conjunction with the power network's macroscopic active power load prediction model. This dynamic response feature vector triggers autonomous operation when the wide area network experiences a power outage. Logically, the end control module uses the global boundary provided by this vector as a feedforward compensation benchmark to correct the real-time status data acquired by the local sensors without delay. This bridges the mechanism gap between the limited field of view of the local control layer and the timing discontinuity of the top-level power grid scheduling. The end control module monitors the reception time difference between two adjacent monitoring messages using a hardware-level clock cycle. When the reception time difference exceeds the preset message reception time threshold, the prediction compensation unit inside the end control module extracts the rotor speed fluctuation envelope and load disturbance amplitude from the last valid message. It then combines this with the operation instructions in the message to complete second-order derivative fitting and determine the displacement increment compensation value of the execution terminal during the communication switching cycle. The end control module uses the dynamic response feature vector in the monitoring message to initiate the kinematic compensation prediction process, calculates and generates a predictive control trajectory that satisfies the mechanical time constant of the gas turbine, and simultaneously determines the upper limit of the rate of change of the predictive control trajectory based on the equivalent inertial constant contained in the dynamic response feature vector. This ensures that the generated virtual control quantity increment is within the physical response envelope of the execution terminal, thereby driving the execution terminal to complete the smooth continuation of physical actions along the autonomous inference trajectory that maintains the second-order continuity of the controlled physical state quantity.

[0030] After the redundant transmission module completes network redirection and distribution to restore the reception of monitoring messages, the terminal control module obtains the control quantity deviation between the predicted control trajectory generated based on autonomous inference and the newly arrived operation command in real time. When the absolute value of the displacement deviation between the predicted control trajectory and the real-time operation command increases, the terminal control module extends the logical approximation period of the execution terminal from the local autonomous prediction mode to the remote real-time control mode by increasing the value of the transition correction coefficient. It uses the transition correction coefficient, which is dynamically updated according to the control quantity deviation, to control the predicted control trajectory to complete the damped convergence to the operation command, driving the execution terminal to absorb the step impact of the command value caused by the link switching and restore it to a stable operating state that follows the command issued by the remote master control terminal. The redundant transmission module monitors the health status of the underlying physical route in real time by receiving the link status announcement message issued by the automatic switching optical network control plane. At the moment when the route protection switching identifier is parsed, the redundant transmission module synchronously sends a high-priority path switching status message to the terminal control module, starts the prediction compensation working mode of the terminal control module, and makes the start time of the autonomous prediction logic and the data interruption time at the physical layer synchronized.

[0031] Example 2: This example constructs a hardware-in-the-loop simulation test bench comprising a digital real-time simulator and an industrial programmable logic controller (PLC). The PLC's sampling frequency is 1kHz. The digital real-time simulator runs the rotor dynamics control equations and the gas turbine thermodynamic state equations, outputting a rotor speed signal superimposed with Gaussian white noise at a signal-to-noise ratio of 25dB to the PLC. This reproduces the electromagnetic interference and measurement disturbance environment. A simulated optical switch path switching action is injected into the communication link to induce timing data interruption. A message reception time threshold is set, determined based on the network's underlying statistical jitter variance and the servo mechanism's mechanical response lag time. When the network's underlying statistical jitter variance decreases and the servo mechanism's response damping increases, the threshold is reduced. To prevent the accumulation of mechanical inertia from causing physical displacement to exceed limits, and considering the asymmetric time delay characteristics generated by network transmission, the end control module extracts the predicted control trajectory's derivation timing scale and the source transmission time stamp added to the operation command after recovery reception in real time. The time difference between the two is mapped to a phase offset vector. The end control module calculates the magnitude of this phase offset vector and multiplies it by a preset physical path transmission hysteresis compensation coefficient. The obtained corrected time parameter is superimposed on the threshold benchmark of the statistical period in the form of positive compensation or reverse reduction to complete the dynamic adjustment of the next statistical period. 10ms is selected as the benchmark communication interaction period, the normal jitter extreme value of the communication plane is measured to be 2ms, and the message reception time threshold is set to 15ms.

[0032] Three gradient conditions were set for communication interruption periods caused by network switching: 20ms, 50ms, and 80ms. Under the baseline condition of a 50ms interruption period, a control group was introduced. This control group froze the output increment after the communication interruption. Test data showed that the gas turbine rotor speed in this control group lacked the outer loop regulation input and deviated from the safety setpoint within the 50ms period. When the network was restored and a new command was issued, the servo execution terminal generated a step opening change of 12%, resulting in a 4.5% overshoot oscillation in the rotor speed frequency. The experimental group used complete command distribution logic, and the end control module monitored the time difference between the reception of two adjacent monitoring messages. When the receiving time difference When it reaches 15ms, among which The physical layer message arrival time interval is represented by the prediction compensation unit inside the end control module, which extracts monitoring messages containing 25dB noise. The prediction compensation unit filters out high-frequency noise components, obtains the net rotor speed angular acceleration, and calculates the dynamic response feature vector by combining it with the instantaneous change rate of load power. The prediction compensation unit uses second-order derivative fitting for the operation command in the last valid message and calculates the displacement increment compensation value of the execution terminal within a 50ms communication switching cycle to be 0.45mm. Based on this displacement increment compensation value, a predictive control trajectory is generated. A partially missing control group is introduced. This partially missing control group uses linear extrapolation for the last operation command, excluding the physical inertia constraint term included in the dynamic response feature vector. Test data shows that the displacement increment calculated by the partially missing control group is 0.62mm, resulting in a 2.1% frequency overshoot when communication is restored. The predictive control trajectory generated by the experimental group is constrained by the upper limit of the equivalent inertia constant in the dynamic response feature vector. The generated virtual control quantity increment matches the gas turbine mechanical time constant. The step opening change during communication restoration is reduced to 0.8%, and the frequency overshoot converges to 0.3%.

[0033] Increasing the communication interruption period to 120ms to test the performance inflection point, the test data shows that after the communication interruption period exceeds 80ms, the predicted control trajectory based on second-order continuity deviates from the actual rotor thermodynamic evolution trajectory, the truncation error of the displacement increment compensation value increases, and the frequency overshoot rises to 3.8%. The test data confirms that 50ms to 80ms is the working range for kinematic compensation prediction using dynamic response characteristic vectors. After a 50ms interruption and resumption of reception, the end control module calculates the absolute value of the displacement deviation between the predicted control trajectory and the real-time issued operation command, and controls the predicted control trajectory to converge to the operation command by dynamically updating the transition correction coefficient, driving the execution terminal to absorb the impact of link switching and restore steady-state operation.

[0034] Example 3: When the industrial control network is in normal operation, the instruction encapsulation module continuously collects the gas turbine rotor speed, load power, and historical displacement parameters of the servo execution terminal at a sampling period of 1ms. The instruction encapsulation module uses a moving average algorithm to smooth the original speed data and determine the net rotor speed. The instruction encapsulation module determines the angular acceleration of the rotor speed by calculating the numerical difference of the net rotor speed in adjacent sampling periods. It extracts the active power signal fed in from the grid side, calculates the power numerical difference in a single sampling period, and determines it as the instantaneous rate of change of the load power. The instruction encapsulation module reads the servo cylinder position data within a set time window, fits the position data sequence using the least squares method, and determines the slope of the fitted line as the motion gradient of the displacement historical parameters. The instruction encapsulation module extracts the angular acceleration, instantaneous rate of change, and motion gradient, and multiplies the angular acceleration by the rotor moment of inertia constant. The instantaneous rate of change is multiplied by the damping power coefficient. The instruction encapsulation module then concatenates the sum of the products of the two products with the motion gradient to generate a dynamic response feature vector. The instruction encapsulation module writes the dynamic response feature vector as a physical inertia constraint term to the end of the data frame of the operation instruction, attaches the source end physical attribute identifier to generate a monitoring message, and sends it to the redundant transmission module. When the communication link is switched over and the end control module resumes receiving monitoring messages, the end control module extracts the current displacement coordinates contained in the locally generated predictive control trajectory and parses the target displacement coordinates contained in the operation instructions in the real-time monitoring message. The end control module calculates the absolute value of the displacement deviation between the current displacement coordinates and the target displacement coordinates. The end control module multiplies the absolute value of the displacement deviation by a preset deviation adjustment slope, accumulates the product result to the initially set reference transition coefficient, and outputs the dynamically updated transition correction coefficient.

[0035] The end-point control module limits the change in the output rate of the predicted control trajectory based on the transition correction coefficient, and controls the output displacement control command to converge towards the target displacement coordinate. The servo execution terminal receives the displacement control command and adjusts the opening of the hydraulic proportional valve to offset the displacement execution deviation caused by the communication path switching, maintaining the steady-state operation of the gas turbine rotor speed. During the offline calibration phase, the commissioning terminal sets the reference transition coefficient. The operation terminal controls the gas turbine to maintain stable no-load operation and disconnects the primary and backup communication links of the redundant transmission module, triggering the end-point control module to start the kinematic compensation prediction process and recording the duration of the communication disconnection. Internal execution terminal predicts displacement increment After restoring the primary and backup communication links, record the time required for the execution terminal to return to steady-state operation under the new operation instructions. Overshoot of actual displacement The memory records multiple sets of different durations. Next corresponding time Overshoot of actual displacement According to the formula Calculate the initial coefficients for each test group, and extract the maximum calculated result from all test groups to determine the baseline transition coefficient. Write to non-volatile memory; in actual operation, when monitoring messages resume normal reception, the end control module extracts the arrival operation command containing the target displacement coordinates. The predicted control trajectory displacement coordinates at this moment The absolute value of the displacement deviation is determined by calculating the absolute value of the difference between the two. Read the internal mapping table using the absolute value of the displacement deviation Matching the corresponding deviation adjustment slope Deviation adjustment slope To achieve a discrete value that increases linearly with the increase of gas turbine load power, with a fixed control period. The step unit is based on the formula Calculate the first Dynamic transition correction coefficient for each cycle , Corresponding to the The absolute value of the displacement deviation in the previous cycle, divided by the displacement control command value in the next cycle and the dynamic transition correction coefficient. The upper limit of the rate of change of the command output drives the hydraulic proportional valve to adjust the opening.

[0036] Example 4: In the field deployment scenario of a gas turbine connected to an automatic switching optical network, the debugging terminal initiates an offline calibration process for the physical inertia of the servo cylinder. A sinusoidal sweep displacement command with a frequency range of 0.1Hz to 10Hz is injected into the hydraulic proportional valve. The actual angular velocity response curve of the rotor is collected, the ratio of the input command amplitude to the peak output angular velocity is calculated, and the attenuation characteristics of the system transfer function are obtained. The debugging terminal defines the mechanical time constant of the gas turbine based on the corner frequency of the transfer function, and determines this mechanical time constant as the initial value of the deviation adjustment slope. Simultaneously, the angular acceleration attenuation trajectory during the gas turbine shutdown test is recorded, and the rotor rotational inertia constant is calculated. The debugging terminal writes the deviation adjustment slope and the rotor rotational inertia constant into the command encapsulation module.

[0037] After the basic physical parameters are solidified, the debugging terminal initiates link switching and joint debugging tests. Pulse truncation interference is injected into the primary fiber optic link to induce monitoring message loss. Within the disconnection window, the end control module initiates a kinematic compensation prediction process, recording the displacement residual between the predicted control trajectory and the newly issued operation command after communication recovery. The debugging terminal calculates the displacement residual expansion rate under increasing disconnection cycles, extracts the maximum allowable residual span that keeps the servo cylinder opening abrupt change rate within a safety margin, divides it by the residual expansion rate to obtain the damping time constant, and uses this damping time constant to set the update step size of the transition correction coefficient. The upper limit limits the trajectory approximation rate of the end control module after network switching recovery, driving the servo loop to converge to a steady-state regulation state after experiencing the command cliff caused by link routing; when the debugging terminal is setting the damping power coefficient, it applies a step power disturbance excitation to the gas turbine in no-load operation, collects the logarithmic reduction rate of the active power signal during the decay period, determines the damping power coefficient of the corresponding unit dynamic characteristics based on the product of the logarithmic reduction rate and the rated speed, and writes the value into the non-volatile storage area of ​​the end control module, limiting the damping compensation intensity of the kinematic compensation prediction process when dealing with load transient jumps.

[0038] Example 5: When a gas turbine is first connected to an automatic optical switching network or restarted after a major overhaul, the commissioning terminal initiates a basic data array filling procedure for the command encapsulation module and the end control module. The commissioning terminal injects a step test signal into the gas turbine controller, acquiring high-frequency sampling points of the rotor's original speed during the free acceleration phase. The command encapsulation module calculates the Nyquist frequency limit of the original speed high-frequency sampling points and defines the initial sliding time window for the moving average algorithm based on this Nyquist frequency limit. While calculating the Nyquist frequency limit, the command encapsulation module performs a fast Fourier transform on the sequence of original speed high-frequency sampling points to extract the highest-order mechanical resonance frequency actually excited by the rotor shaft system. The frequency component is used as the upper limit of the effective signal frequency and substituted into the sampling theorem to establish the Nyquist limit point. Through this frequency mapping logic, the limit is based on the connotation of physical mechanical dynamic characteristics, thereby ensuring that the defined sliding time window length can physically fully cover and smooth the high-frequency mechanical disturbance cycle. The end control module receives the benchmark test message with the initial sliding time window identifier, calculates the spectral dispersion of the rotor speed fluctuation envelope under different window scales, compares the spectral dispersion with the inherent mechanical resonance frequency of the gas turbine, and extracts the sliding time window value corresponding to the minimum dispersion that avoids the resonance frequency band, and solidifies it as the reference scale for obtaining the net rotor speed during normal operation.

[0039] After fixing the sliding time window value, the debugging terminal simulates a communication interruption delay increasing from 0ms to the network switching extreme value within the local area network, causing the end-control module to trigger the kinematic compensation prediction process. The prediction compensation unit within the end-control module extracts the operation instructions from the benchmark test message and constructs a third-order state matrix encompassing position, velocity, and acceleration parameters. During the matrix construction process, the prediction compensation unit directly uses the target opening encapsulated in the operation instructions as the system position parameter and simultaneously calls the dynamic response feature vector of the last valid message, directly converting its contained motion gradient data into velocity parameters; simultaneously, it extracts the angular acceleration from the feature vector and... The acceleration parameter is obtained by multiplying the equivalent transmission ratio coefficient of the execution terminal. This enables the complete completion of all state matrix dimension variables with only the composite information of a single message. The prediction compensation unit inputs the third-order state matrix into the second-order derivative fitting operator to calculate the fitting residual between the polynomial approximation curve and the actual rotor speed trajectory. The debugging terminal extracts the polynomial coefficient distribution characteristics corresponding to the minimum fitting residual and maps them as the weight benchmark for calculating the displacement increment compensation value. This completes the calculation constraints inside the prediction compensation unit. The end control module uses this weight benchmark to generate a predictive control trajectory and drives the execution terminal to eliminate the displacement abrupt change at the beginning of the path switching along a smooth curve that matches the current mechanical state.

[0040] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0041] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A command distribution logic control system for a gas turbine monitoring environment, characterized in that, The system includes: The instruction encapsulation module is used to obtain the gas turbine rotor speed, load power, and historical displacement parameters of the execution terminal, and to determine the dynamic response feature vector based on the angular acceleration of the rotor speed, the instantaneous rate of change of the load power, and the motion gradient of the historical displacement parameters. The instruction encapsulation module binds the dynamic response feature vector as a physical inertia constraint to the operation instruction in a time sequence and encapsulates it into a monitoring message with source physical attribute identifier. The redundant transmission module, connected to the instruction encapsulation module, is used to build a primary and backup dual-plane communication link based on the automatic switching optical network. Within a 50ms switching period after the primary link failure is detected, the monitoring message is redirected to the backup link for distribution through the routing protection algorithm. The end-point control module, connected to the redundant transmission module, monitors the time difference between two consecutive monitoring messages. When the time difference exceeds a preset message reception time threshold, the end-point control module initiates a kinematic compensation prediction process using the dynamic response feature vector within the monitoring message. It calculates and generates a predictive control trajectory that satisfies the mechanical time constant of the gas turbine, driving the execution terminal to complete a smooth displacement. After normal reception of monitoring messages resumes, it extracts the target displacement coordinates from the arriving operation command. The predicted control trajectory displacement coordinates at this moment The absolute value of the displacement deviation is determined by calculating the absolute value of the difference between the two. Read the internal mapping table using the absolute value of the displacement deviation Matching the corresponding deviation adjustment slope Deviation adjustment slope To achieve a discrete value that increases linearly with the increase of gas turbine load power, with a fixed control period. The step unit is based on the formula Calculate the first Dynamic transition correction coefficient for each cycle ,in, As the benchmark transition coefficient, Corresponding to the The absolute value of the displacement deviation in the previous cycle, divided by the displacement control command value in the next cycle and the dynamic transition correction coefficient. The upper limit of the rate of change of the command output drives the hydraulic proportional valve to adjust the opening, and controls the predicted control trajectory to complete the damped convergence of the operation command.

2. The instruction distribution logic control system for a gas turbine monitoring environment according to claim 1, characterized in that, The end control module is equipped with a prediction compensation unit. The prediction compensation unit extracts the rotor speed fluctuation envelope and load disturbance amplitude from the dynamic response feature vector in real time, and combines it with the operation command in the last valid message to complete the second derivative fitting, and determines the displacement increment compensation value of the execution terminal during the communication switching vacuum period, so as to maintain the second-order continuity of the controlled physical state quantity.

3. The instruction distribution logic control system for a gas turbine monitoring environment according to claim 1, characterized in that, When the redundant transmission module initiates the link protection switching logic, it synchronously sends a path switching status message to the end control module. After receiving the path switching status message, the end control module reduces the message reception time threshold to improve the end control module's sensitivity to wide area network transmission jitter and packet loss events.

4. The instruction distribution logic control system for a gas turbine monitoring environment according to claim 1, characterized in that, The dynamic response characteristic vector includes the equivalent inertia constant of the execution terminal; the end control module determines the upper limit of the rate of change of the predicted control trajectory based on the equivalent inertia constant, ensuring that the increment of the generated virtual control quantity does not exceed the physical response envelope of the execution terminal, and avoiding nonlinear oscillations in the gas turbine combustion chamber induced by command jumps.

5. A command distribution logic control system for a gas turbine monitoring environment according to claim 1, characterized in that, If the absolute value of the displacement deviation between the predicted control trajectory and the real-time issued operation command increases, the end control module increases the value of the transition correction coefficient to extend the logical approximation period of the execution terminal from the local autonomous prediction mode to the remote real-time control mode, thereby offsetting the step impact of the command value caused by the link switching.

6. A command distribution logic control system for a gas turbine monitoring environment according to claim 1, characterized in that, The redundant transmission module determines message continuity by verifying the cyclic sequence count value in the header of the monitoring message protocol; If the cumulative transmission duration corresponding to the number of interruptions in the cyclic sequence count exceeds the message reception time threshold, the end control module automatically switches to the autonomous logic processing state based on the dynamic response feature vector.

7. A command distribution logic control system for a gas turbine monitoring environment according to claim 1, characterized in that, The dynamic response feature vector is associated with the load power envelope, which characterizes the thermodynamic features of the gas turbine. The terminal control module performs dynamic gain correction on the predicted control trajectory based on the load power envelope to offset the rotor speed frequency drift caused by the transient load on the grid side, thus ensuring the physical self-consistency of the gas turbine control logic.

8. A command distribution logic control system for a gas turbine monitoring environment according to claim 1, characterized in that, The end control module calculates the phase offset vector between the predicted control trajectory and the restored operation command in real time, and adjusts the message reception time threshold of the next statistical period according to the phase offset vector to adaptively match the asymmetric transmission delay differences caused by different physical path switching.

9. A command distribution logic control system for a gas turbine monitoring environment according to claim 1, characterized in that, After completing the communication path switching action, the redundant transmission module sends a register reset signal to the end control module; the end control module responds to the register reset signal, clears the prediction cache stack established based on the dynamic response feature vector, and re-completes the synchronization reference calibration of the system logic clock.

Images