Hydraulic turbine guide vane opening control system based on cloud control architecture

Abstract

The application discloses a guide vane opening control system of a hydraulic turbine unit based on a cloud control architecture, which comprises a data center, a master cloud controller, a slave cloud controller, a PID controller and a switching execution unit. The data center generates an operation data frame with a collection time mark and distributes the operation data frame to each controller; each controller calculates a respective control output based on the operation data frame; the master cloud controller and the slave cloud controller transmit the internal states of the controllers; the switching execution unit selects a target control output based on the control outputs of the controllers and the internal states of the controllers, executes output maintenance, deviation compensation and ramp transition on the target control output, and generates a guide vane opening control instruction and outputs the guide vane opening control instruction to a guide vane execution mechanism. The application places the controller redundancy switching and the PID / GPC control strategy switching in a unified data and state consistency framework, and reduces the guide vane opening mutation in the switching process.

Description

Technical Field

[0001] This application relates to the field of hydropower technology, specifically to a guide vane opening control system for a hydro turbine unit based on a cloud-based control architecture. Background Technology

[0002] The guide vane opening of a hydro turbine unit is a key actuator for regulating the unit's influent flow and output power. The accuracy and stability of guide vane opening control directly affect the unit's load response performance and operational safety. Currently, guide vane opening control in hydro turbine units generally adopts a single-machine PID (Proportional-Integral-Derivative) control architecture. A local governor or programmable logic controller calculates the control quantity based on load commands and guide vane opening feedback, and directly outputs it to the guide vane actuator.

[0003] The aforementioned control architecture has at least the following shortcomings in actual operation: First, the fixed-parameter PID controller has limited adaptability to head changes, load disturbances, and turbine nonlinearity, making it difficult to simultaneously guarantee steady-state accuracy and dynamic response speed over a wide operating range. Second, the single-controller architecture lacks redundancy. When the controller fails, requires maintenance, or undergoes firmware upgrades, the control output is prone to interruption or abrupt changes, leading to guide vane opening fluctuations and unit load oscillations. Third, in scenarios where optimized control algorithms need to be introduced to improve control performance, there is usually a lack of internal state connection mechanism between the optimized controller and the original PID controller, making it easy to cause secondary disturbances when directly switching between the two control algorithms. Fourth, the redundancy switching at the controller level and the strategy switching at the control algorithm level are usually designed independently, with each maintaining its own data source and state information, making it difficult to complete continuous takeover under a unified data consistency and state consistency framework.

[0004] Therefore, there is an urgent need for a turbine guide vane opening control system that can achieve controller redundancy and coordinated switching of control strategies based on unified data. Summary of the Invention

[0005] The purpose of this application is to provide a turbine guide vane opening control system based on a cloud-based control architecture to solve the problems of insufficient adaptability of the existing single-machine PID control architecture, independent switching of controller redundancy and control strategy, and lack of a unified data and state consistency framework.

[0006] To achieve the above objectives, the technical solution adopted in this application is as follows: This application provides a cloud-based control architecture for a turbine guide vane opening control system, including a data platform, a main cloud controller, a slave cloud controller, a PID controller, and a switching execution unit. The data platform is connected to the main cloud controller, the slave cloud controller, and the PID controller. Both the main cloud controller and the slave cloud controller include a generalized predictive control module, and the PID controller includes a PID control module. The main cloud controller, the slave cloud controller, and the PID controller are connected to the switching execution unit. The main cloud controller is connected to the slave cloud controller. The switching execution unit is connected to the guide vane actuator. The data platform is used to generate and distribute running data frames with acquisition time stamps; the main cloud controller, slave cloud controller, and PID controller are used to calculate and output the main control output, slave control output, and PID control output based on the running data frames, respectively; the main cloud controller is also used to transmit the internal state of the controller to the slave cloud controller and the switching execution unit; the switching execution unit is used to select the target control output based on the main control output, slave control output, PID control output, and the internal state of the controller, and after performing output holding, deviation compensation, and ramp transition on the target control output, it generates the guide vane opening control command and outputs it to the guide vane actuator.

[0007] Optionally, a business data channel, a state synchronization channel, and a heartbeat monitoring channel are provided between the master cloud controller and the slave cloud controller; the internal state of the controller transmitted by the state synchronization channel includes model state variables, control parameters, and internal controller variables.

[0008] Furthermore, the switching execution unit includes a first switching execution unit. When the first switching execution unit detects a fault in the main cloud controller, a communication abnormality detected by the heartbeat monitoring channel, or receives a controller switching command, it compares the output deviation between the main control output and the slave control output, and verifies the synchronization integrity of the controller's internal state and the consistency of the acquisition time stamp of the running data frame. When the preset master-slave switching conditions are met, the first switching execution unit selects the slave control output as the target control output.

[0009] Furthermore, when the first switching execution unit selects the control output as the target control output, it sequentially performs output hold, deviation compensation, and ramp transition on the control output to form the guide vane opening control command.

[0010] Optionally, the switching execution unit includes a second switching execution unit; the currently active controller and the non-currently active controller calculate their respective control outputs in parallel based on the running data frame; wherein, the currently active controller includes any one of the master cloud controller, the slave cloud controller, and the PID controller, and the non-currently active controller includes the remaining controllers among the master cloud controller, the slave cloud controller, and the PID controller excluding the currently active controller; the second switching execution unit determines the non-currently active controller as the controller to be switched, and makes the controller to be switched track the control output and internal state of the currently active controller; when the preset control strategy switching conditions are met, the second switching execution unit selects the control output of the controller to be switched to take over the guide vane control.

[0011] Furthermore, the control strategy switching conditions include stable system operation, normal communication, output deviation between PID control output and main control output or slave control output meeting preset requirements, and receiving a valid control strategy switching command; the controller to be switched tracks the currently active controller through output tracking, deviation compensation, and internal state synchronization.

[0012] Optionally, the generalized predictive control module includes a model identification submodule. The model identification submodule extracts dynamic response data of guide vane opening and unit power under different load conditions from historical operating data, establishes a first-order inertial plus pure time delay model based on the dynamic response data, and uses particle swarm optimization algorithm to determine the gain, inertial time, and delay time of the first-order inertial plus pure time delay model.

[0013] Furthermore, the generalized predictive control module calculates the master control output or slave control output based on a first-order inertial plus pure time delay model and the prediction time domain, control time domain, guide vane opening constraint, and guide vane opening change rate constraint.

[0014] Optionally, the data platform is equipped with a data acquisition interface. The data platform obtains unit load commands, guide vane opening feedback, unit power and control status through the data acquisition interface, adds acquisition time stamps to form running data frames, and synchronously distributes the running data frames to the main cloud controller, the slave cloud controller and the PID controller.

[0015] Optionally, the switching execution unit outputs a single control output to the guide vane actuator only within the same control cycle, and the switching execution unit performs output arbitration based on the controller activity status flag and the switching enable flag.

[0016] Compared with the prior art, this application has the following advantages: This application generates unified operational data frames with acquisition time stamps through a data platform and simultaneously distributes them to the master cloud controller, slave cloud controller, and PID controller, enabling the three types of controllers to perform parallel calculations based on the same data benchmark. Through internal state transmission between the master cloud controller and slave cloud controller and joint verification by the switching execution unit, continuous takeover between the master and slave controllers is achieved. Through the tracking management and transition processing of PID control output and generalized predictive control output by the switching execution unit, bumpless switching between PID control and generalized predictive control is achieved. This places controller redundancy switching and control strategy switching under a unified data and state consistency framework, reducing abrupt changes in guide vane opening during the switching process. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall architecture of a cloud-based control architecture-based guide vane opening control system for a hydro-turbine unit, according to an embodiment of this application. Figure 2 This is a schematic diagram of the channel connection between a master cloud controller and a slave cloud controller according to an embodiment of this application; Figure 3 This is a schematic diagram of the composition of a switching execution unit according to an embodiment of this application. Detailed Implementation

[0018] The present application will now be further described with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present application and should not be construed as limiting the scope of protection of the present application. It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present application.

[0019] The acquisition, transmission, storage, use, and processing of data in this application comply with relevant national laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0020] All terms used in this application have the same meaning as understood by one of ordinary skill in the art to which this application pertains, unless otherwise specifically defined. It should also be understood that terms defined in general dictionaries should be interpreted as having meanings consistent with their meanings in the context of the relevant art, and not as idealized or highly formalized, unless expressly defined herein.

[0021] It should be noted that the cloud-based controller in this application refers to a software-defined controller deployed in a virtualized resource pool or cloud computing environment. It receives operational data and outputs control commands via network communication, rather than a traditional local hardware speed controller. Generalized Predictive Control (GPC) is an advanced control algorithm based on a controlled object model that continuously optimizes the control quantity in the prediction time domain. A PID controller refers to a controller that uses a proportional-integral-derivative control law and can be deployed in the original simulation system or local control equipment.

[0022] like Figure 1 As shown, the turbine guide vane opening control system 1 provided in this embodiment includes a data platform 10, a main cloud controller 20, a slave cloud controller 30, a PID controller 40, and a switching execution unit 50. The data platform 10 has a data acquisition interface, which acquires the turbine load command, guide vane opening feedback, unit power, and control status variables, and adds acquisition timestamps to each of these data to form an operating data frame. The main cloud controller 20, slave cloud controller 30, and PID controller 40 are connected to the data platform 10 to receive the same operating data frame. Both the main cloud controller 20 and slave cloud controller 30 include a generalized predictive control module, which performs generalized predictive control calculations based on the operating data frame to obtain the main control output and slave control output, respectively. The PID controller 40 includes a PID control module, which performs PID control calculations based on the operating data frame to obtain the PID control output. The main cloud controller 20, the slave cloud controller 30, and the PID controller 40 are connected to the switching execution unit 50 to input the main control output, slave control output, and PID control output to the switching execution unit 50. The main cloud controller 20 is connected to the slave cloud controller 30 to transmit the internal state of the controller. The switching execution unit 50 is connected to the guide vane actuator 2, selects the target control output based on the main control output, slave control output, PID control output, and the internal state of the controller, and performs output hold, deviation compensation, and ramp transition on the target control output to generate a guide vane opening control command, which is then output to the guide vane actuator 2.

[0023] It should be noted that the PID controller and its corresponding PID control module are conventional and mature technologies used in existing turbine guide vane control, and are not improvements or innovations of this application. The core improvement of this application lies in: adding a cloud-based control architecture based on generalized predictive control, and designing a switching execution unit capable of seamless dual-mode switching, thus constructing a dual-control redundancy switching system of the original conventional PID control and the newly added GPC optimized control. This application implements optimized control logic through the main cloud controller and the GPC control module built into the slave cloud controller, and relies on the switching execution unit to complete the flexible and bumpless switching between the traditional PID control strategy and the new GPC control strategy, thereby improving the dynamic adjustment performance and operational fault tolerance of the unit while retaining the original mature control methods. Therefore, the technical contribution of this application is limited to the GPC cloud control architecture, model identification and optimization method, and dual-mode control switching mechanism, and does not involve any improvement to the existing PID control itself.

[0024] Furthermore, it should be noted that the data platform 10 in this embodiment supports multi-scenario data interaction and communication adaptation, adapting to both simulation debugging scenarios and actual engineering operation scenarios. Specifically, in simulation debugging scenarios, the data platform can interact with the simulator system using the OPC UA (Open Platform Communications Unified Architecture) protocol. An OPC UA server node is deployed on the simulator side, and the data platform 10 acts as a client, subscribing to operating parameters, status variables, and control command data. In actual engineering scenarios, the data platform can directly connect to the actual unit's measurement and control equipment on-site, collecting real operating data, operating status, and control feedback data without relying on the simulator system, thus meeting the dual application needs of simulation debugging and actual on-site operation. Simultaneously, the data platform 10 can transmit operating data frames with the main cloud controller 20 and the slave cloud controller 30 using power communication protocols (such as IEC 104). The selection of the above communication methods falls within the scope of engineering implementation and does not constitute a limitation on the scope of protection of this application.

[0025] In one possible implementation, the data platform 10 adds acquisition timestamps to load commands, guide vane opening feedback, unit power, and control status variables to form operational data frames. The acquisition timestamps are generated using a unified clock source to ensure comparability of the timestamps for each data item. Each operational data frame contains three fields: data item identifier, data value, and acquisition timestamp. The data platform 10 sends the operational data frames to the main cloud controller 20, the slave cloud controller 30, and the PID controller 40, ensuring that the three types of controllers receive identical input data in each control cycle. The purpose of this implementation is to provide a unified data benchmark for subsequent parallel computing and switching verification, avoiding control output deviations caused by inconsistent data sources or misaligned timestamps.

[0026] In one possible implementation, such as Figure 2 As shown, the master cloud controller 20 and the slave cloud controller 30 are connected by a service data channel 21, a state synchronization channel 22, and a heartbeat monitoring channel 23. The service data channel 21 is used for control data interaction between the master cloud controller 20 and the slave cloud controller 30, including the forwarding and confirmation of running data frames and the interaction of control outputs. The state synchronization channel 22 is used to transmit the internal state of the controller, which includes model state variables, control parameters, and controller internal variables. Model state variables refer to the internal state vector of the controlled object model in the generalized predictive control module; control parameters refer to algorithm parameters such as prediction time domain, control time domain, and control weighting coefficients; and controller internal variables refer to intermediate variables and historical control quantity sequences during the rolling optimization process. The heartbeat monitoring channel 23 is used for the periodic exchange of heartbeat messages between the master cloud controller 20 and the slave cloud controller 30 to monitor the operational health status and communication link status of the other party.

[0027] While the master cloud controller 20 is in active control mode, the slave cloud controller 30 continuously receives model state variables, control parameters, and internal controller variables from the master cloud controller 20 via the state synchronization channel 22, and updates its own internal state to match that of the master cloud controller 20. Simultaneously, the slave cloud controller 30 performs deviation compensation on the slave control output based on the master control output, ensuring that the deviation between the master and slave control outputs remains within a preset deviation threshold range. The purpose of this implementation is to ensure that the slave cloud controller 30 is always in a hot standby state, enabling it to directly output control quantities continuous with the master cloud controller 20 when takeover is required, avoiding output jumps caused by inconsistencies in internal states.

[0028] In one possible implementation, such as Figure 3 As shown, the switching execution unit 50 includes a first switching execution unit 51 and a second switching execution unit 52. The first switching execution unit 51 is used for switching between the master cloud controller 20 and the slave cloud controller 30, and the second switching execution unit 52 is used for switching between the original PID controller 40 and the cloud controller (master cloud controller 20 or slave cloud controller 30).

[0029] In one possible implementation, when the first switching execution unit 51 detects a fault in the main cloud controller 20, a communication anomaly detected by the heartbeat monitoring channel 23, or receives a controller switching command, it executes the following switching determination process: First, it compares the output deviation between the main control output and the slave control output to determine whether the output deviation is less than or equal to a preset deviation threshold; second, it verifies the synchronization integrity of the controller's internal state, determining whether the model state variables, control parameters, and controller internal variables transmitted by the state synchronization channel 22 are complete and updated within the effective time window; third, it verifies the consistency of the acquisition timescale of the running data frames, determining whether the running data frames currently used by the main cloud controller 20 and the slave cloud controller 30 have the same acquisition timescale. When the output deviation is less than or equal to the preset deviation threshold, the controller's internal state is synchronized completely, and the acquisition timescale is consistent, the first switching execution unit 51 determines that the preset master-slave switching conditions are met, selects the slave control output as the target control output, and the slave cloud controller 30 takes over the guide vane control.

[0030] This implementation uses output deviation, internal state synchronization integrity, and data timescale consistency as joint verification items. Its purpose is to ensure that when the cloud controller 30 takes over, not only is the output value close to that of the main cloud controller 20, but the internal state and input data are also completely consistent, thereby avoiding the gradual deviation of the control quantity due to state divergence after takeover.

[0031] In one possible implementation, when the first switching execution unit 51 selects the slave control output as the target control output, it sequentially performs output holding, deviation compensation, amplitude limiting, and ramp transition on the slave control output. Output holding means that in the first control cycle after the switch, the first switching execution unit 51 keeps the guide vane opening control command unchanged before the switch, so that the guide vane actuator 2 receives continuous control quantities at the moment of switching. Deviation compensation means that the switching execution unit applies a compensation amount to the slave control output according to the difference between the master control output and the slave control output at the time of switching, so that the compensated slave control output is connected with the master control output before the switch. Amplitude limiting means that the compensated control quantity is limited to the upper and lower limits of the allowable guide vane opening, preventing abnormal output from causing the guide vane to open or close too much. Ramp transition means that in several control cycles after the switch, the guide vane opening control command gradually transitions from the value before the switch to the target output value of the cloud controller 30 according to a preset ramp rate, avoiding guide vane opening oscillation caused by sudden changes in control quantity. The purpose of this implementation is to minimize the disturbance of the master-slave controller switching on the guide vane opening through multi-level transition processing.

[0032] In one possible implementation, the currently active controller and non-currently active controllers calculate their respective control outputs in parallel based on the same runtime data frame. The currently active controller includes any one of the master cloud controller 20, the slave cloud controller 30, and the PID controller 40. The non-currently active controllers include the remaining controllers among the master cloud controller, slave cloud controller, and PID controller excluding the currently active controller. The second switching execution unit 52 determines the non-currently active controller as the controller to be switched and makes the controller to be switched track the control output and internal state of the currently active controller. Specifically, when the PID controller 40 is the currently active controller, the master cloud controller 20 (or the slave cloud controller 30) acts as the controller to be switched. Through an output tracking stage, the generalized predictive control output tracks the PID control output. Through a deviation compensation stage, the steady-state deviation between the two is eliminated. Through an internal state synchronization stage, the integral state and other internal variables of the PID controller are mapped to the initial state of the generalized predictive control module. When the master cloud controller 20 (or the slave cloud controller 30) is the currently active controller, the PID controller 40 acts as the controller to be switched. Through a similar tracking and synchronization mechanism, the PID control output tracks the generalized predictive control output.

[0033] When the preset control strategy switching conditions are met, the second switching execution unit 52 takes over the guide vane control using the control output of the controller to be switched. The control strategy switching conditions include stable system operation, normal communication, the output deviation between the PID control output and the generalized predictive control output meeting preset requirements, and receiving a valid control strategy switching command. Stable system operation means that both the unit power deviation and the guide vane opening deviation are within a stable threshold range for a preset time; normal communication means that the communication link between the data center 10 and each controller is uninterrupted and the data timestamp is valid; the output deviation meeting preset requirements means that the difference between the PID control output and the generalized predictive control output is less than or equal to the preset strategy switching deviation threshold; a valid control strategy switching command is a switching command from the human-machine interface or the upper-level scheduling system, and the command has a valid timestamp and checksum.

[0034] The purpose of this implementation is to enable PID control and generalized predictive control to run in parallel and track each other on the same running data frame, and to achieve continuous takeover when the control strategy switching conditions are met, thereby supporting the safe commissioning of generalized predictive control and reliable rollback in abnormal situations.

[0035] In one possible implementation, the generalized predictive control module includes a model identification submodule. The model identification submodule extracts dynamic response data of guide vane opening and unit power under different load conditions from historical operating data. Specifically, this includes recording the unit power response curve after a step change in guide vane opening under each load condition, and extracting characteristic parameters such as steady-state gain, rise time, settling time, and overshoot from the response curve. Based on the above dynamic response data, the model identification submodule abstracts the controlled object into a first-order inertial plus pure time-delay model, whose transfer function is as follows: Where K is the gain, T is the inertia time, and τ is the delay time. The model identification submodule uses Particle Swarm Optimization (PSO) to determine the above model parameters, with the sum of squared errors between the simulation output and the actual running data as the objective function. The optimal solution is searched in the parameter space through iterative evolution of the particle swarm. The purpose of this implementation is to provide a high-precision predictive model for generalized predictive control, overcoming the limitation of difficult-to-accurate parameter acquisition in traditional mechanistic modeling.

[0036] In one possible implementation, the generalized predictive control module calculates the generalized predictive control output based on a first-order inertial plus pure time-delay model, as well as the prediction time domain, control time domain, guide vane opening constraint, and guide vane opening rate of change constraint. The prediction time domain refers to the time span over which the controller predicts future outputs, while the control time domain refers to the number of control steps optimized by the controller in each control cycle. The guide vane opening constraint defines the upper and lower limits of the guide vane opening, and the guide vane opening rate of change constraint defines the upper limit of the change in guide vane opening within adjacent control cycles. Within each control cycle, the generalized predictive control module predicts the unit power response in the future prediction time domain based on the current operating data frame and the controlled object model. Under the condition of satisfying the guide vane opening constraint and the rate of change constraint, it optimizes the control quantity sequence in the control time domain using the weighted sum of the load command tracking error and the change in control quantity as the objective function, and takes the first element of this sequence as the generalized predictive control output for the current control cycle. The purpose of this implementation is to achieve forward-looking adjustment of the guide vane opening through rolling optimization, thereby improving control accuracy and dynamic response speed under wide operating conditions.

[0037] In one possible implementation, the switching execution unit 50 outputs a single control output to the guide vane actuator 2 only within the same control cycle. The switching execution unit 50 maintains a controller activity status flag and a switching enable flag. The controller activity status flag identifies which controller (master cloud controller 20, slave cloud controller 30, or PID controller 40) is currently active, and the switching enable flag indicates whether controller switching or control strategy switching is permitted. The switching execution unit 50 selects the corresponding control output as the current output based on the controller activity status flag and updates the controller activity status flag when the switching enable flag is valid and the switching conditions are met, thus completing output arbitration. The purpose of this implementation is to ensure that the guide vane actuator 2 receives only one defined control quantity in each control cycle, avoiding control conflicts caused by multiple controllers outputting simultaneously.

[0038] For example, in one specific implementation, a turbine unit has a rated capacity of 300MW, a guide vane opening adjustment range of 0% to 100%, and a control cycle of 50ms. The data platform 10 collects load commands, guide vane opening feedback, unit power, and control status variables every 50ms. After adding a collection time stamp, it forms an operating data frame and distributes it to the main cloud controller 20, the slave cloud controller 30, and the PID controller 40. The generalized predictive control modules in the main cloud controller 20 and the slave cloud controller 30 adopt a first-order inertial plus pure time delay model. The model gain K=0.85, inertial time T=6.21s, and delay time τ=3s are obtained through particle swarm optimization algorithm. The prediction time domain start time of the generalized predictive control is set to 3s, the prediction time domain length is set to 18 control steps, the control time domain is set to 5 control steps, and the guide vane opening change rate is constrained to no more than 0.5% per control cycle. The master cloud controller 20 and the slave cloud controller 30 interact through a business data channel, a status synchronization channel, and a heartbeat monitoring channel. The status synchronization cycle is one control cycle, and the heartbeat monitoring cycle is 100ms. The preset deviation threshold is set to 0.3% of the guide vane opening, and the ramp transition rate is set to 0.1% per control cycle. When the master cloud controller 20 fails, the switching execution unit 50, after detecting the loss of heartbeat, jointly verifies the output deviation (measured 0.15%, less than 0.3%), the integrity of internal status synchronization (the most recent synchronization was within 50ms, complete), and the consistency of the acquisition timescale (master and slave use the same acquisition timescale to run data frames). If the master-slave switching conditions are met, the slave control output is selected as the target control output. After output holding, deviation compensation, amplitude limiting, and ramp transition, the output is sent to the guide vane actuator 2. During the switching process, the maximum fluctuation of the guide vane opening is less than 0.5%, and the unit power fluctuation is less than 0.2% of the rated capacity.

[0039] To further improve the system's engineering availability and on-site operation and maintenance efficiency, this system is equipped with an independent and unified human-machine interface (HMI). As an external, independent device, the HMI establishes a two-way communication link with the system's backend database, enabling data interaction and operational linkage. The HMI can read various operational information stored in the backend database in real time, including real-time unit operating data, controller operating status, generalized predictive control (GPC) parameters, control strategy switching flags, and system anomaly alarm information. It also supports online issuance of operation commands by operators, allowing them to write control activation / deactivation commands, parameter modification values, control mode selections, and other operational data into the backend database in real time. This data is then read, parsed, and transmitted to each controller and switching execution unit by the high-performance data platform to update and execute the corresponding control logic. Through this HMI, operators can intuitively and conveniently complete all system operation and maintenance tasks, including switching between PID control and GPC control strategies, managing redundancy switching of master-slave cloud controllers, maintaining control model parameters, adjusting control time-domain and predictive time-domain parameters, setting upper and lower limit thresholds for guide vane opening control, and real-time monitoring of the entire system's operating status. Meanwhile, the human-machine interface can visually display the changing trends of key process variables of the unit, the switching status of controllers, the status of system communication connections, and real-time abnormal alarm information, realizing integrated management of control strategy deployment, logic switching, status monitoring, and equipment operation and maintenance. This human-machine interface design effectively simplifies the on-site operation process of the system, significantly improves the visualization, maintainability, and on-site application value of the cloud-based control system, and ensures the stable, efficient, and convenient operation of the unit's guide vane opening control system.

[0040] It should be noted that the terms "first," "second," and similar terms used in this application do not indicate any order, quantity, or importance, but are merely used to distinguish different parts. Terms such as "including" or "comprising" mean that the element preceding the word covers the element listed after the word, and do not exclude the possibility of covering other elements as well.

[0041] Although the operations are described in a specific order in the accompanying drawings in this application, it should not be construed as requiring these operations to be performed in the specific order or serial order shown, or requiring all of the shown operations to obtain the desired result. In certain environments, multitasking and parallel processing may be advantageous.

[0042] Finally, it should be noted that the above content is only used to illustrate the technical solution of this application, and is not intended to limit the scope of protection of this application. Simple modifications or equivalent substitutions made by those skilled in the art to the technical solution of this application shall not depart from the substance and scope of the technical solution of this application.

Claims

1. A guide vane opening control system for a hydro-turbine unit based on a cloud-based control architecture, characterized in that, This includes a data platform, a main cloud controller, a slave cloud controller, a PID controller, and a switching execution unit; The data platform is connected to the main cloud controller, the slave cloud controller, and the PID controller, respectively; both the main cloud controller and the slave cloud controller include a generalized predictive control module, and the PID controller includes a PID control module; the main cloud controller, the slave cloud controller, and the PID controller are respectively connected to the switching execution unit; the main cloud controller is connected to the slave cloud controller; the switching execution unit is connected to the guide vane actuator. The data platform is used to generate and distribute running data frames with acquisition timestamps; the main cloud controller, the slave cloud controller, and the PID controller are respectively used to calculate and output the main control output, slave control output, and PID control output based on the running data frames; the main cloud controller is also used to transmit the internal state of the controller to the slave cloud controller and the switching execution unit; the switching execution unit is used to select a target control output based on the main control output, the slave control output, the PID control output, and the internal state of the controller, and after performing output holding, deviation compensation, and ramp transition on the target control output, generate a guide vane opening control command and output it to the guide vane actuator.

2. The turbine guide vane opening control system based on cloud-based control architecture according to claim 1, characterized in that, The master cloud controller and the slave cloud controller are provided with a business data channel, a status synchronization channel and a heartbeat monitoring channel; the internal status of the controller transmitted by the status synchronization channel includes model state variables, control parameters and internal controller variables.

3. The turbine guide vane opening control system based on a cloud-based control architecture according to claim 2, characterized in that, The switching execution unit includes a first switching execution unit. When the first switching execution unit detects a fault in the main cloud controller, a communication abnormality detected by the heartbeat monitoring channel, or receives a controller switching command, it compares the output deviation between the main control output and the slave control output, and verifies the synchronization integrity of the controller's internal state and the consistency of the acquisition timescale of the running data frame. When the preset master-slave switching conditions are met, the first switching execution unit selects the slave control output as the target control output.

4. The turbine guide vane opening control system based on cloud-based control architecture according to claim 3, characterized in that, When the first switching execution unit selects the slave control output as the target control output, it sequentially performs output hold, deviation compensation, and ramp transition on the slave control output to form the guide vane opening control command.

5. The turbine guide vane opening control system based on cloud-based control architecture according to claim 1, characterized in that, The switching execution unit includes a second switching execution unit; the currently active controller and the non-currently active controller calculate their respective control outputs in parallel based on the running data frame; wherein, the currently active controller includes any one of the master cloud controller, the slave cloud controller, and the PID controller, and the non-currently active controller includes the remaining controllers among the master cloud controller, the slave cloud controller, and the PID controller excluding the currently active controller; the second switching execution unit determines the non-currently active controller as the controller to be switched, and makes the controller to be switched track the control output of the currently active controller and the internal state of the controller; when the preset control strategy switching conditions are met, the second switching execution unit selects the control output of the controller to be switched to take over the guide vane control.

6. The turbine guide vane opening control system based on cloud-based control architecture according to claim 5, characterized in that, The control strategy switching conditions include stable system operation, normal communication, output deviation between the PID control output and the main control output or the slave control output meeting preset requirements, and receiving a valid control strategy switching command; the controller to be switched tracks the currently active controller through output tracking, deviation compensation, and internal state synchronization.

7. The turbine guide vane opening control system based on cloud-based control architecture according to claim 1, characterized in that, The generalized predictive control module includes a model identification submodule. The model identification submodule extracts dynamic response data of guide vane opening and unit power under different load conditions from historical operating data, establishes a first-order inertial plus pure time delay model based on the dynamic response data, and uses particle swarm optimization algorithm to determine the gain, inertial time, and delay time of the first-order inertial plus pure time delay model.

8. The turbine guide vane opening control system based on cloud-based control architecture according to claim 7, characterized in that, The generalized predictive control module calculates the master control output or the slave control output based on the first-order inertial plus pure time delay model, as well as the prediction time domain, control time domain, guide vane opening constraint, and guide vane opening change rate constraint.

9. The turbine guide vane opening control system based on cloud-based control architecture according to claim 1, characterized in that, The data platform is equipped with a data acquisition interface. The data platform acquires unit load commands, guide vane opening feedback, unit power and control status variables through the data acquisition interface, adds the acquisition time stamp to form the running data frame, and synchronously distributes the running data frame to the main cloud controller, the slave cloud controller and the PID controller.

10. The turbine guide vane opening control system based on cloud-based control architecture according to claim 1, characterized in that, The switching execution unit outputs a single control output to the guide vane actuator within the same control cycle, and the switching execution unit performs output arbitration based on the controller activity status flag and the switching enable flag.

Images