An offline testing method and platform for flight simulator control load systems

Abstract

This invention belongs to the field of flight simulation equipment testing, and relates to an offline testing method and platform for flight simulator control load systems. It aims to solve the problems of ECL component testing being dependent on the entire machine, inefficient, and unable to independently diagnose. The platform includes a transmission mechanism composed of a joystick and a rack and pinion, and is equipped with a control system consisting of a controller and a servo driver, and a drive assembly for automatic testing. The method includes: comparing the actual motion position with the target trajectory command to generate a main drive control component; simultaneously comparing the actual output torque with the target torque curve to generate a torque compensation adjustment component; and superimposing the torque compensation adjustment component onto the main drive control component to form a comprehensive control signal, thereby achieving high-precision force-position composite simulation. This invention enables independent, efficient, and accurate offline testing of ECL components, significantly shortening maintenance cycles, reducing costs, and improving the depth and reliability of testing.

Description

Technical Field

[0001] This invention belongs to the field of flight simulation equipment testing, and specifically relates to an offline testing method and platform for the control load system of a flight simulator. Background Technology

[0002] Flight simulators are crucial equipment for training pilots. Through highly realistic simulation environments, they enable pilots to master flight skills and handle various complex situations. Within a flight simulator, the control load system is one of the core subsystems, and its force feedback function is primarily achieved through the Electronic Control Loading (ECL). This component provides precise force feedback to the pilot's control devices, such as the control stick and pedals, to simulate the aerodynamic loads of a real aircraft under different flight conditions, and is key to achieving a high-fidelity flight control feel.

[0003] However, in existing technologies, performance testing, calibration, and fault diagnosis of ECL components typically require installation on a complete flight simulator for integrated testing. This testing method has significant drawbacks: First, the testing process occupies the entire simulator, resulting in long downtime, severely impacting normal training schedules and reducing equipment availability; second, the testing process is complex, inefficient, and makes it difficult to accurately pinpoint faults to the ECL component itself, increasing maintenance difficulty and time costs; third, due to its reliance on the entire simulator environment, independent and comprehensive functional verification cannot be performed after component repair or before spare parts are stocked, posing a risk of introducing potential faults into the system. Therefore, the industry urgently needs a dedicated solution that can independently, efficiently, and accurately test ECL components without requiring a simulator environment. Summary of the Invention

[0004] To address the aforementioned problems in the prior art, namely the technical issues that ECL component testing in the prior art relies on the entire machine, is inefficient, and cannot be independently diagnosed, the present invention provides an offline testing method and platform for flight simulator control load systems.

[0005] In a first aspect, the present invention provides an offline testing method for a flight simulator control load system, the method comprising the following steps:

[0006] The electronically controlled payload assembly under test is mounted on the test platform;

[0007] Based on the set simulated flight state parameters, generate target motion trajectory commands and target torque curve commands;

[0008] Real-time acquisition of the actual motion position feedback and actual output torque feedback of the electronically controlled load assembly;

[0009] The actual motion position feedback is compared with the target motion trajectory command, which is the main control target, to determine the position error, and the main drive control component is generated based on the position error.

[0010] The actual output torque feedback is compared with the target torque curve command, which serves as an auxiliary correction target, to determine the torque error, and a torque compensation adjustment component is generated based on the torque error.

[0011] The torque compensation adjustment component is superimposed on the main drive control component to form a comprehensive adjustment control signal;

[0012] The test platform's drive components analyze and execute the integrated adjustment and control signals to drive the mechanical transmission mechanism connected to the electronically controlled load assembly.

[0013] Furthermore, the method also includes a transmission delay measurement step, specifically including:

[0014] In response to the input step motion command, record the first time point at which the step motion command was issued;

[0015] Continuously monitor the actual motion position feedback;

[0016] When the value of the actual movement position feedback is detected to reach the preset trigger threshold for the first time, the corresponding second time point is recorded;

[0017] The motion response delay time of the electronically controlled load assembly is obtained by calculating the difference between the second time point and the first time point.

[0018] Furthermore, the method includes a manual testing mode, in which:

[0019] Provides a manually operated joystick connected to the mechanical transmission mechanism;

[0020] The use of the target motion trajectory command is stopped, and the real-time displacement of the joystick is used as the input of the main drive control component, while the generation and superposition of the torque compensation adjustment component are maintained.

[0021] During the manual operation of the joystick, the displacement data of the joystick and the actual torque data output by the electronic control load component are recorded simultaneously to form a displacement-torque relationship curve.

[0022] The displacement-torque relationship curve is compared and analyzed with a pre-stored standard curve conforming to aviation standards.

[0023] Furthermore, the method includes an automated mechanical performance testing mode, in which:

[0024] Receive input automated test parameters, which include at least the motion distance and the number of reciprocating motions;

[0025] Set the target torque curve command to zero or a constant value, and control the drive component to drive the component to reciprocate according to the automated test parameters;

[0026] During the reciprocating motion, the continuous output of the position sensor of the electronically controlled load assembly is collected and recorded to generate a position-time curve;

[0027] Analyze the smoothness and linearity of the position-time curve.

[0028] Furthermore, the method also includes a comprehensive fault diagnosis step based on multi-source information cross-validation, specifically including:

[0029] The system simultaneously acquires position sensor signals and force sensor signals from the electronically controlled load assembly, as well as encoder signals from the drive component.

[0030] The acquired signals are compared and correlated with the driving commands issued to the driving components in real time.

[0031] Based on the predetermined matching rules and error tolerances that characterize the inherent physical correlation between signals when the system is working normally, determine whether the logical and numerical relationships between signals are within the normal range.

[0032] When it is determined that at least one logical relationship or numerical relationship exceeds the normal range, the system is deemed to have a functional fault, and a shutdown and alarm are triggered.

[0033] Furthermore, prior to the execution of the method, a step of parameter tuning is included for the control algorithm used to generate the main drive control component and the torque compensation adjustment component. This step is achieved by performing a series of step response tests.

[0034] Input the step-like target motion trajectory command into the system;

[0035] Record and analyze the time response curve of the actual motion position feedback, and extract its dynamic performance indicators such as overshoot, rise time and settling time;

[0036] Based on the dynamic performance index, the internal parameters of the control algorithm are iteratively adjusted until the dynamic performance index meets the preset performance requirements.

[0037] In a second aspect, the present invention provides an offline test platform for a flight simulator control load system, used for testing electronic control load components, the platform comprising:

[0038] The mechanical transmission mechanism includes a control lever, a drive motor, a first transmission component, a second transmission component, and a mechanical output end;

[0039] The joystick is driven to the input end of the first transmission component;

[0040] The output end of the first transmission component is connected to the second transmission component and can drive the second transmission component as a whole to perform linear motion;

[0041] The drive motor is connected to the second transmission component in a driving connection;

[0042] The mechanical output end is located on the output end of the second transmission component and is used to connect the output link of the electronically controlled load component.

[0043] Drive and control systems, including servo drives and controllers;

[0044] The servo driver is electrically connected to the drive motor;

[0045] The controller is connected to the position sensor and force sensor of the servo driver and the electronically controlled load assembly, respectively.

[0046] The controller is used to receive feedback signals from the position sensor and force sensor, and generate drive commands according to test requirements, and control the action of the drive motor through the servo driver.

[0047] Furthermore, the first transmission assembly includes a first gear and a first rack;

[0048] The first gear is connected to the control lever, the first rack meshes with the first gear, and the first rack is connected to the second transmission assembly.

[0049] Furthermore, the second transmission assembly includes a housing, a second gear, and a second rack;

[0050] The drive motor is mounted on the housing, and its output shaft is connected to the second gear transmission.

[0051] The second rack meshes with the second gear, and the second rack is fixedly connected to the mechanical output end, which can move linearly relative to the housing;

[0052] At least a portion of the housing is connected to the first rack, such that the linear motion of the first rack can drive the housing and the second transmission assembly to move as a whole.

[0053] Furthermore, the platform also includes a mounting mechanism, which includes a base plate and at least one clamp assembly;

[0054] The substrate is provided with a positioning structure;

[0055] The clamp assembly is detachably connected to the positioning structure and can be adjusted along the positioning structure to fix electronically controlled load components of different specifications.

[0056] The beneficial effects of this invention are:

[0057] This invention constructs an independent offline testing platform, enabling the testing, calibration, and fault diagnosis of the Electronic Control Load (ECL) component to be performed entirely outside the flight simulator. This directly avoids prolonged simulator occupancy due to testing and maintenance, significantly reducing equipment downtime and thus substantially increasing the effective training time and availability of the flight simulator, while reducing indirect costs caused by equipment downtime.

[0058] The platform of this invention integrates automated testing processes and dedicated diagnostic software, enabling standardized durability testing, transmission delay measurement, and performance curve comparison. Compared to manual debugging in complex system environments, the testing process of this invention is faster and more repeatable. Furthermore, its multi-source information cross-validation fault diagnosis method can quickly and accurately pinpoint faults to specific modules of the ECL component, significantly improving maintenance efficiency.

[0059] This invention employs industrial-grade servo drives, high-precision encoders, and closed-loop control algorithms to accurately reproduce and control the motion trajectory and torque output of ECL components. Through PID regulation and software compensation mechanisms, the system's control accuracy and stability are effectively improved. This ensures the objectivity and accuracy of test results, guaranteeing that repaired or newly purchased ECL components meet stringent performance standards before installation.

[0060] Because this invention possesses independent offline testing capabilities, the maintenance department can conduct comprehensive quality verification on ECL components before they are put into storage or installed on the machine. This proactive maintenance strategy can effectively screen out spare parts with potential defects, preventing faulty components from entering the operating system at the source, thereby improving the reliability and operational safety of the entire flight simulator system.

[0061] The mechanical mounting mechanism of this invention features adjustable clamps and positioning structures, making it compatible with ECL components of different specifications and models, thus exhibiting excellent versatility. Simultaneously, the integrated software system provides clear manual / automatic testing mode switching and a graphical data display interface, simplifying the operation process, reducing the skill requirements for testing personnel, and enabling users to complete the entire testing task conveniently and quickly. Attached Figure Description

[0062] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0063] Figure 1 This is a flowchart of an offline testing method for a flight simulator control load system according to the present invention;

[0064] Figure 2 This is a schematic diagram of the structure of an offline test platform for a flight simulator control load system according to the present invention;

[0065] Figure 3 yes Figure 2 A magnified view of a portion of the image. Detailed Implementation

[0066] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0067] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0068] The first embodiment of the present invention proposes an offline testing method for a flight simulator control load system, the method comprising the following steps:

[0069] Step S10: Install the electronic control load assembly under test onto the test platform;

[0070] Step S20: Based on the set simulated flight state parameters, generate target motion trajectory command and target torque curve command;

[0071] Step S30: Real-time acquisition of the actual motion position feedback and actual output torque feedback of the electronically controlled load component;

[0072] Step S40: Compare the actual motion position feedback with the target motion trajectory command, which is the main control target, to determine the position error, and generate the main drive control component based on the position error;

[0073] Step S50: Compare the actual output torque feedback with the target torque curve command, which serves as an auxiliary correction target, to determine the torque error, and generate a torque compensation adjustment component based on the torque error;

[0074] Step S60: The torque compensation adjustment component is superimposed on the main drive control component to form a comprehensive adjustment control signal;

[0075] Step S70: The integrated adjustment control signal is analyzed and executed by the drive component of the test platform to drive the mechanical transmission mechanism connected to the electronically controlled load assembly.

[0076] To more clearly illustrate the offline testing method for a flight simulator control load system according to the present invention, the following is in conjunction with... Figure 1 The steps in the embodiments of the present invention are described in detail below:

[0077] Step S10: Install the electronic control load assembly under test onto the test platform;

[0078] First, the Electronically Controlled Load Cell (ECL) under test is mounted and secured onto the mechanical transmission mechanism of the test platform. Then, its motion output terminal is mechanically connected to the platform's transmission mechanism to transmit motion and force. Finally, the electrical interfaces for control and signal feedback of the ECL component are connected to the test platform's control system.

[0079] This completes the physical and electrical integration of the component under test and the test platform, preparing for subsequent offline testing procedures.

[0080] Step S20: Based on the set simulated flight state parameters, generate target motion trajectory command and target torque curve command;

[0081] Step S20 in the offline testing method of this invention is essentially about converting the discrete and macroscopic simulated flight state parameters set by the operator into continuous and precise control commands that the testing platform can execute. In this embodiment, this step is automatically completed by the host computer or embedded main controller of the testing platform. Specifically, the control system pre-stores a core multi-dimensional aerodynamic data model, which can exist in the form of a database, multi-dimensional lookup table, or complex function fitting, and it completely describes the aerodynamic characteristics of a specific aircraft in different flight envelopes.

[0082] When the operator inputs simulated flight parameters through the human-machine interface, such as a flight speed of Mach 0.8, a flight altitude of 30,000 feet, an angle of attack of 2 degrees, and flap retraction status, the main controller initiates the command generation program. The program first uses these input flight parameters as index coordinates, performing lookup and interpolation calculations in a preset aerodynamic data model. Through algorithms such as multidimensional linear interpolation, it accurately calculates the hinge torque magnitude experienced by the aircraft control surfaces, such as ailerons and elevators, at any angle of rotation under that specific flight condition. This calculation process does not generate a single torque value, but rather a complete curve describing the functional relationship between control surface rotation angle and hinge torque.

[0083] Subsequently, the system invokes another built-in transmission mechanism and artificial sensory model, the construction of which is itself a key technical prerequisite for the high-fidelity reproduction of the manipulation sensation in this invention. The construction of this model can be achieved through, but is not limited to, two technical paths:

[0084] The first approach is analytical modeling based on aircraft design drawings and the mechanical principles of the flight control system. By analyzing the geometric relationships and mechanical properties of all mechanical and hydraulic components, such as the control stick, lever, rocker arm, torque tube, and even hydraulic booster, technicians establish a complete set of kinematic mapping relationships that can accurately describe the kinematic mapping relationship from the control stick displacement to the control surface deflection angle, as well as the mathematical equations for the force-displacement-velocity characteristics generated by the sensory system composed of spring assemblies, dampers, counterweights, and servo valves.

[0085] The second, more empirical approach is system identification and experimental data fitting. Technicians can conduct calibration experiments on real aircraft, high-fidelity ground integrated test platforms (so-called iron birds), or full-motion flight simulators. By applying preset displacements and velocities at the control inputs and simultaneously measuring the displacement, velocity, and required force of the control stick using high-precision displacement and force sensors, as well as measuring the actual deflection angle of the control surfaces at the output, a large number of input-output data pairs can be collected. Then, system identification algorithms such as polynomial fitting, neural network training, or least squares methods can be used to reverse-engineer the model parameters that best reproduce the measured data, or a high-dimensional lookup table can be directly generated with control stick displacement and velocity as inputs and control stick force as outputs. Linear or spline interpolation methods can be used during the lookup to ensure smooth and continuous output.

[0086] The system utilizes this precisely constructed model to accurately transform the hinge torque-control surface angle curve obtained in the previous step into a control stick force-control stick displacement relationship curve that the pilot can actually feel in the cockpit. This transformed and mapped curve is the final target torque curve command. Simultaneously, to drive the movement of the component under test to simulate the pilot's control actions, the system generates a time-series position command based on the selected test condition, such as simulating a uniform stick pull, a sharp roll, or a periodic sinusoidal control input. This command defines the precise position the control stick should reach every millisecond during the test, thus forming a time-displacement function curve, which is the target motion trajectory command.

[0087] For example, in a test simulating a constant-speed pull to the bottom of the stick within 2 seconds, the target motion trajectory command is a straight line that increases uniformly from zero displacement to maximum travel. Ultimately, the controller packages these two related but independently generated commands in a time-synchronized manner. The target motion trajectory command specifies the sequence of component positions on the time axis, while the target torque curve command defines the required torque value to be applied at each position point on the motion trajectory. This packaged high-density data stream is then prepared to be sent to the servo drive system, enabling precise closed-loop control of the electronic control payload components' position and force feedback, thus reproducing the control feel under specified flight conditions with high fidelity.

[0088] Step S30: Real-time acquisition of the actual motion position feedback and actual output torque feedback of the electronically controlled load component;

[0089] Step S30 in the offline testing method of this invention has the core task of constructing a high-precision, high-frequency feedback link to provide the necessary real-time status data for the subsequent closed-loop control algorithm. In specific implementation, this step is accomplished through a sensor system integrated within or tightly coupled to the electronically controlled load assembly, along with corresponding data acquisition hardware. For the feedback acquisition of the actual motion position, the system uses a high-resolution absolute rotary encoder or resolver. This sensor is coaxially mounted at the end of the output shaft of the servo motor. Alternatively, to eliminate errors introduced by backlash and elastic deformation in the transmission chain, it is more preferably directly coupled to the output link at the very end of the electronically controlled load assembly, thereby directly measuring the actual angular or linear displacement applied to the test object. The encoder can instantly output a digital signal representing the current absolute position, such as Gray code or a serial synchronization signal. This signal is directly decoded by a dedicated interface circuit within the servo driver or main controller, eliminating the need for power-on zeroing and ensuring the immediacy and accuracy of the position data.

[0090] For the feedback acquisition of the actual output torque, a high-stiffness, high-sensitivity strain gauge torque sensor is installed in series between the servo motor output shaft and the transmission chain of the control mechanism. The physical position of this sensor is crucial, ensuring that the measured torque is the total torque actually output by the system to the external test object, including the motor's own output torque and all dynamic effects during transmission. Since the electrical signal generated by the strain gauge during deformation is extremely weak, typically in the millivolt range and susceptible to electromagnetic interference, a dedicated signal conditioning module is configured adjacent to the sensor. This module includes a high common-mode rejection ratio instrumentation amplifier for precise signal amplification and a low-pass filter to remove high-frequency noise. The conditioned analog voltage signal is then fed into a high-precision analog-to-digital converter (ADC), such as a 24-bit Σ-Δ ADC, to convert it into a digital quantity with extremely high resolution. These two independent sensor data streams—digitized position information and digitized torque information—are sampled at high speed in a strictly synchronized manner by the embedded main controller or a dedicated data acquisition card.

[0091] In this embodiment, the sampling frequency is set to no less than 1000 Hz to ensure that any rapid dynamic changes during operation can be captured and to provide sufficient time margin for the control loop. For each synchronous sampling, the controller assigns a unified, high-resolution timestamp to the pair of position and torque data, thereby generating two time-synchronized, discrete data sequences. These two data sequences together constitute a complete real-time feedback signal, providing accurate state information for subsequent calculation of control errors and generation of drive commands.

[0092] Step S40: Compare the actual motion position feedback with the target motion trajectory command, which is the main control target, to determine the position error, and generate the main drive control component based on the position error;

[0093] Step S40 in the offline testing method of this invention is the core link in realizing high-precision position servo tracking in the digital control domain, which is essentially a real-time running position closed-loop control law. Specifically, in each control cycle, typically within a millisecond-level time interval synchronized with the data acquisition in step S30, the embedded main controller first reads the actual motion position feedback value provided by the sensor system and retrieves the target motion trajectory instruction value corresponding to the same timestamp from the memory. The controller then performs a subtraction operation, that is, subtracts the actual position from the target position, thereby instantaneously calculating the deviation between the two, which is the position error signal. This position error signal is not directly used as the control output, but as the input of the classic closed-loop control algorithm. In this embodiment, the system uses a precisely tuned proportional-integral-derivative (PID) controller to generate the main drive control component based on the position error.

[0094] Specifically, the proportional element in the PID controller responds instantly to the current position error, generating a corrective torque proportional to the error magnitude. This torque serves as the primary power source for driving the load to rapidly approach the target position. The integral element, by accumulating all position errors over a period of time, eliminates steady-state errors caused by factors such as friction and gravitational imbalance, ensuring that the electronically controlled load components can ultimately stop precisely at the target position without deviation, rather than stopping near the target. The derivative element provides a damping effect by calculating the rate of change of the position error, i.e., predicting the future trend of the error. Its purpose is to suppress overshoot and oscillations that may occur during rapid movement, improving the dynamic response stability and smoothness of the system. In each control cycle, the main controller weights and sums these three components according to preset gain coefficients Kp, Ki, and Kd obtained through system identification or empirical tuning. The final value obtained is the main drive control component. This component is a digital instruction that directly represents the theoretical driving torque required by the servo motor to overcome inertia and friction and drive the load to follow the target trajectory with optimal dynamic characteristics. This main drive control component will then be further processed and superimposed with the torque compensation component generated in subsequent steps to form the final complete drive instruction issued to the servo driver.

[0095] Step S50: Compare the actual output torque feedback with the target torque curve command, which serves as an auxiliary correction target, to determine the torque error, and generate a torque compensation adjustment component based on the torque error;

[0096] In step S50 of the offline testing method of this invention, a torque servo adjustment loop is constructed in the entire control system. This loop operates in parallel with the position control loop described in step S40. Its core purpose is not to directly generate motion, but to make real-time and precise corrections to the main drive control components, thereby ensuring that the actual torque applied by the electronically manipulated load assembly to the test object can reproduce the target torque curve defined in step S20 with high fidelity, thus creating an accurate and expected maneuvering force feel for the tester.

[0097] In practice, within each control cycle, the controller first needs to determine the appropriate target torque value based on the current physical state of the system. Since the target torque curve is essentially a function mapping the manipulation position to the output torque, the controller directly references the actual motion position feedback value acquired in step S30 of the current cycle, using it as a key input. The use of the actual position rather than the target position here is based on a core control consideration: the force feedback should precisely correspond to the current actual position of the manipulation device to provide the operator with a stable and physically consistent interactive experience. The controller uses this real-time position value to index the target torque curve data structure, i.e., a lookup table, pre-loaded into memory. Since the real-time position value rarely falls exactly on discrete data points in the lookup table, to obtain a smooth and continuous target torque, the controller performs a fast interpolation calculation, such as using linear interpolation or a higher-order spline interpolation algorithm, to calculate the precise instantaneous target torque value between two adjacent data points.

[0098] The controller subtracts the instantaneous target torque value, which changes dynamically with position and is obtained through interpolation, from the actual output torque feedback value collected in step S30 during the same period. The result is the torque error. This error signal is a quantified deviation value that precisely reveals the difference between the actual output torque and the expected output torque at the current physical position. This torque error signal is then sent to a regulation algorithm module, designed specifically for torque adjustment and independent of the position controller.

[0099] In this embodiment, the module employs a fast-responding proportional-integral (PI) controller. The proportional element instantaneously generates a proportional correction based on the current torque error magnitude, rapidly narrowing the gap between the actual and target torque. Simultaneously, the integral element accumulates torque errors over a period of time, generating a correction that eliminates steady-state torque deviations, effectively overcoming persistent errors caused by inaccurate models, variations in transmission system friction, or sensor zero-point drift.

[0100] Finally, the controller performs a weighted sum of the proportional and integral components output by the PI regulator to obtain a digital adjustment command, which is the torque compensation adjustment component. Physically, this component represents the additional increment or decrement required to compensate for the error between the actual torque and the target torque in the main drive command of the motor. As a feedforward compensation signal, it actively corrects the driving force calculated by the main controller based on an ideal model. This allows the entire servo system to not only accurately track position commands but also precisely shape and render every detail of the target torque curve, providing a highly realistic control force feedback environment for the entire offline testing process.

[0101] Step S60: The torque compensation adjustment component is superimposed on the main drive control component to form a comprehensive adjustment control signal;

[0102] Step S60 in the offline testing method of this invention performs arithmetic merging of control signals, which is the final output stage for realizing the dual closed-loop control strategy. In the calculation process of each control cycle, this step performs digital addition on the main drive control component output from the position control loop in step S40 and the torque compensation adjustment component output from the torque servo loop in step S50, thereby generating a single comprehensive adjustment control signal.

[0103] In practice, the central processing unit within the controller sums two independent digital quantities—the main drive control component and the torque compensation adjustment component—as operands within a synchronous calculation cycle. From a control engineering perspective, the main drive control component is calculated based on position error. Its main function is to drive the actuator to generate macroscopic motion to overcome system inertia and basic damping, thereby tracking the predetermined target trajectory. The torque compensation adjustment component, on the other hand, is calculated based on real-time torque error. Its function is to generate a correction signal to accurately compensate for various mechanical nonlinearities and disturbances not fully covered by the system model. These factors include, but are not limited to, motor cogging torque, nonlinear friction of the transmission system, unforeseen external load fluctuations, and zero-point drift of the torque sensor.

[0104] By performing an addition operation, the torque compensation adjustment component is directly superimposed on the main drive control component, resulting in a comprehensive adjustment control signal that functionally satisfies the requirements of two control dimensions simultaneously. It includes both the basic drive quantity required to achieve the position tracking target and the precise compensation quantity required to achieve the torque reproduction target. Technically, this control structure constitutes an effective fusion of feedforward compensation and feedback control, enabling the torque loop output to directly correct the drive command ultimately applied to the servo motor.

[0105] The integrated control signal, as the final output of the digital controller within this control cycle, is transmitted to the input of the power amplifier stage of the servo drive system. Depending on the system hardware configuration, this digital signal may need to be processed by a digital-to-analog converter (DAC) to become an analog voltage signal before transmission, or it may be encapsulated and transmitted as a data frame using a standardized industrial fieldbus protocol, such as CANopen or EtherCAT. After receiving this integrated control signal, the servo amplifier precisely modulates the current output to the servo motor windings according to its value. Ultimately, the physical output torque generated by the motor will be precisely proportional to the value of the integrated control signal. In this way, the dual control objectives of high-fidelity real-time regulation of output torque are achieved throughout the entire process of accurately tracking the motion trajectory.

[0106] Step S70: The integrated adjustment control signal is analyzed and executed by the drive component of the test platform to drive the mechanical transmission mechanism connected to the electronically controlled load assembly.

[0107] Step S70 in the offline testing method of this invention is the physical execution stage of the control command. In this step, the integrated adjustment control signal generated in the previous steps is converted into actual mechanical motion and force by the drive system of the test platform.

[0108] In practice, the drive component is typically a servo driver that first receives a comprehensive adjustment and control signal from the main controller. This signal contains the complete torque command that ultimately needs to be applied to the load.

[0109] Next, the servo driver parses the instruction and precisely adjusts the electrical energy supplied to the servo motor accordingly. This process is achieved by controlling the current supplied to the motor windings, as the motor's output torque is directly related to this current value. In this way, the servo driver efficiently converts an electrical signal instruction into a precisely measured physical torque on the servo motor's output shaft.

[0110] Finally, this physical torque is transmitted through a mechanical transmission mechanism connected to the motor, the end of which is coupled to the electronically controlled load assembly. The end result is that, driven by the torque generated by the motor, the electronically controlled load assembly not only completes the movement to the target position, but also precisely replicates the control command requirements in the applied force, thus achieving closed-loop control of the entire testing method at the physical level.

[0111] In some preferred embodiments, the method further includes a transmission delay measurement step, specifically comprising:

[0112] In response to the input step motion command, record the first time point at which the step motion command was issued;

[0113] Continuously monitor the actual motion position feedback;

[0114] When the value of the actual movement position feedback is detected to reach the preset trigger threshold for the first time, the corresponding second time point is recorded;

[0115] The motion response delay time of the electronically controlled load assembly is obtained by calculating the difference between the second time point and the first time point.

[0116] The primary technical objective of this step is to accurately quantify the inherent time delay of the entire control and execution chain. This delay is a key dynamic parameter of the system, encompassing the entire process from software instruction generation to the mechanical system's initial response, including controller calculation delay, data bus transmission delay, servo drive response delay, and the start-up delay of the motor's electromagnetic and mechanical inertia. Measuring this parameter is fundamental for high-precision system modeling and controller parameter tuning.

[0117] The specific implementation method of this measurement step is as follows:

[0118] First, when the system is stationary, the control program generates and issues a step motion command. The key feature of this command is a sudden, instantaneous change in the target position setpoint within a calculation cycle, for example, instantly changing from the current position P0 to the target position P1. The amplitude of this command, i.e., the difference between P1 and P0, should be set large enough to ensure the response signal significantly exceeds the sensor's noise floor, but not so large that the system enters the nonlinear saturation region or triggers hardware protection. Internally, when the step motion command is formally placed into the data buffer to be sent or marked as ready for execution, the controller immediately accesses its built-in high-resolution hardware clock or timer to read and record the current timestamp, which is defined as the first time point, denoted as t1. The accuracy of this time point recording directly affects the accuracy of the measurement results; therefore, a system clock with microsecond-level or higher precision is typically used.

[0119] After issuing the command, the control system continuously monitors the actual motion position feedback. This feedback signal originates from a high-precision position sensor mounted on the moving parts of the mechanical system, such as an absolute encoder coaxially mounted with a servo motor or a linear encoder mounted on the final actuator. The controller continuously samples the sensor's output signal at its inherent high-speed sampling frequency, such as several kilohertz (kHz). Each sample yields a digital quantity representing the current physical position of the mechanical mechanism.

[0120] During continuous monitoring, the program compares each newly sampled position value with a preset trigger threshold in real time. Setting this trigger threshold is a crucial technical aspect. It is typically set as a small percentage of the total amplitude of the step command, such as one to five percent. Its value must be greater than the noise fluctuation range of the position feedback signal when the system is stationary to prevent false triggering caused by random noise; simultaneously, this value must be small enough to ensure it accurately captures the starting point of the physical motion, rather than a later stage in the response process. When the absolute value of the monitored actual motion position feedback value relative to the initial position P0 first reaches or exceeds this preset trigger threshold, the program determines that the mechanical system has begun to respond effectively to the command. At the instant this condition is detected, the controller accesses its high-resolution hardware clock again and records the corresponding current timestamp, which is defined as the second time point, denoted as t2.

[0121] Finally, the controller performs a simple arithmetic subtraction, subtracting the first time point t1 from the second time point t2. The resulting difference, Δt = t2 - t1, is the motion response delay time of the electronically controlled load component to be measured. This delay time value serves as an important system identifier parameter and will be used for subsequent control algorithm optimization. For example, when designing a feedforward controller, this delay time can be used to pre-compensate for control commands, thereby significantly reducing the system's tracking error. Simultaneously, it is also an important basis for tuning the derivative and integral terms of feedback controllers, such as PID controllers, helping to maximize the system's response speed while ensuring system stability.

[0122] In this embodiment, the parameter tuning step further includes an adaptive learning and online parameter tuning process, which specifically includes:

[0123] While inputting the step-like target motion trajectory command into the system to excite the electronically controlled payload component to generate a motion response, the time-domain response data of the actual motion position feedback and the actual output torque feedback are simultaneously collected.

[0124] Based on the time-domain response data, the dynamic model of the electronic control load component is identified online and updated in real time through the built-in recursive system identification algorithm. The dynamic model uses equivalent moment of inertia, equivalent damping coefficient and equivalent stiffness as key characterization parameters.

[0125] Extract the dynamic response features of the online updated dynamic model, and calculate the matching degree between the features and the ideal dynamic response features corresponding to the preset performance benchmark model; the dynamic response features include at least the overshoot, rise time and settling time in the unit step response curve generated by the model simulation, or the amplitude frequency and phase frequency characteristics of the model at a specific frequency point;

[0126] Based on the feature deviation calculated according to the matching degree, a set of control parameter adjustment amounts are dynamically calculated and output through the parameter mapper;

[0127] The control parameter adjustment amount is applied to the control algorithm in real time to adjust the internal parameters on which the main drive control component and the torque compensation adjustment component are based when generating them online, so that the subsequent step response test is performed under the updated control parameters.

[0128] In one specific embodiment of the present invention, the adaptive learning and online parameter tuning process included in the parameter tuning step is implemented in the following details. When the system inputs a step-like target motion trajectory command to the electronically controlled payload component under test, the command is numerically represented by a sudden change in the position setpoint within a single control cycle, for example, an instantaneous jump from an initial value of 0 mm to a target value of 50 mm. This sudden change command is used to excite the component to generate a dynamic response. During this process, the data acquisition system synchronously records two key time-domain feedback signals at a fixed high sampling frequency fs (e.g., fs = 10 kHz, indicating the acquisition of 10,000 data points per second): one is the actual motion position feedback sequence from the position sensor, which can be denoted as a function x of time t. actual(t) The other path is the actual output torque feedback sequence obtained from the force sensor or by converting the servo motor current and torque constant, denoted as τ. actual(t) The length of the data collection time window T acq It needs to sufficiently cover dynamic processes (e.g., T) acq =0.5 seconds), thus obtaining the complete dataset {t, x actual(t) , τ actual(t) This provides a data foundation for subsequent model identification.

[0129] After obtaining the time-domain data, the system uses a built-in recursive system identification algorithm to perform online identification and real-time updates of the component's dynamic model. This algorithm employs a recursive least squares method with a forgetting factor. The system's preset parameterized model is a linear second-order system model, with torque as input and position as output. Its transfer function in the continuous time domain can be expressed as G(s) = 1 / (Js). 2+Bs+K). In this formula, s is the Laplace operator in the complex frequency domain, J represents the equivalent moment of inertia to be identified, which physically represents the inertial resistance of the component and load to acceleration; B represents the equivalent damping coefficient to be identified, reflecting the viscous friction characteristics during motion; and K represents the equivalent stiffness to be identified, characterizing the elasticity of the system's resistance to positional displacement. This continuous model, after discretization (e.g., using the zero-order hold method, assuming a sampling period Ts=1 / fs), is transformed into a difference equation form suitable for digital computation.

[0130] At each discrete sampling time k, the recursive least squares algorithm performs the following recursive update steps: First, construct the data vector φ(k) and the output scalar y(k) based on the latest input and output data. Next, calculate the gain vector L(k) = P(k-1) × φ(k) / (λ + φ(k)). T ×P(k-1)×φ(k)). Here, P(k) is a covariance matrix reflecting the uncertainty of the current parameter estimate; T denotes matrix transpose; λ is the forgetting factor, a positive number slightly less than 1 (typically 0.98), which assigns higher weight to new data, enabling the algorithm to track time-varying characteristics. Then, the parameter estimate vector θ(k) = θ(k-1) + L(k)×[y(k) - φ(k)] is updated. T [×θ(k-1)], where the vector θ(k) contains the estimates of J, B, and K at the current time. Finally, the covariance matrix P(k) is updated to P(k) = (IL(k) × φ(k)). T )×P(k-1) / λ, where I is the identity matrix. With each step test, newly acquired data is continuously input into this recursive formula, thereby continuously refreshing the estimated values ​​of parameters J, B, and K, and completing the online update of the dynamic model.

[0131] Subsequently, the system changes from the currently updated dynamic model G. current(s) Quantifiable dynamic response features are extracted. In the time domain, the response curve of the model to a unit step input (i.e., a step signal with an amplitude of 1) is calculated through numerical simulation, and characteristic indicators are extracted from this curve: overshoot OS, defined as the percentage by which the peak response exceeds the steady-state value; rise time Tr, typically defined as the time required for the response to rise from 10% to 90% of the steady-state value; and settling time Ts, defined as the time required for the response to enter and remain within ±2% (or other specified error band) of the steady-state value. In the frequency domain, the frequency response of the model at specific frequency points ωc (e.g., ωc1 = 10 rad / s and ωc2 = 30 rad / s) is calculated, including the amplitude M(ωc) (the ratio of the output amplitude to the input sinusoidal signal) and the phase φ(ωc) (the phase shift of the output relative to the input). These extracted features are compared with a preset performance benchmark model G. desired(s)(For example, the ideal eigenvalues ​​of an ideal second-order system with a specific damping ratio, such as 0.707, and no overshoot are compared. The comparison process quantifies the matching degree by calculating the feature deviation. For example, a time-domain deviation vector E is constructed.) time =[OS current -OS desired Tr current -Tr desired Ts current -Ts desired ];

[0132] Among them, OS current This represents the actual overshoot of the tested Electronic Control Load Unit (ECL) and its control system in a step response. It is a percentage value, calculated as (peak value - steady-state value) / steady-state value × 100%. It quantifies the maximum overshoot magnitude by which the system response first exceeds the target steady-state value.

[0133] OS desired This represents a preset, ideal target value for overshoot. It is set based on a performance benchmark model (such as a well-damped second-order system) or practical application requirements (such as certain aerospace standards that may require no overshoot). Its value is typically set to 0% (no overshoot) or a very small positive percentage (allowing for a small amount of overshoot).

[0134] Tr current This represents the actual rise time of the current system response. It is typically defined as the time it takes for the system output to rise from 10% of the initial steady-state value of the step response to 90% of the final steady-state value. It measures the initial response speed of the system to a command.

[0135] Tr desired This represents a preset, ideal target value for the ascent time. It depends on the required speed of the system; for example, a shorter ascent time might be set to simulate agile flight control.

[0136] Ts current This represents the actual settling time of the current system response. It is usually defined as the shortest time required for the system output to go from a step response to entering and remaining within a specified error band (e.g., ±2%) of its final steady-state value. It measures how quickly the system eliminates transient fluctuations and eventually stabilizes.

[0137] Ts desired This represents the preset, ideal settling time target value. It reflects the requirements for the overall stability and smoothness of the system.

[0138] Ultimately, a comprehensive scalar deviation measure can be calculated, such as the weighted Euclidean norm D. time =sqrt(w1×E time [1]2 +w2×E time [2] 2 +w3×E time [3] 2 ), where w1, w2, and w3 are preset weighting coefficients used to balance the importance of different feature indicators.

[0139] Based on the calculated characteristic deviation (such as D) time The parameter mapper dynamically calculates and outputs a set of control parameter adjustments. The parameter mapper can be a rule-based expert system. For example, a rule could be set: if the OS... current Significantly larger than OS desired The output command then increases the differential gain Kd in the control algorithm, because the differential action helps suppress overshoot; if Tr current Significantly greater than Tr desired If the output command increases the proportional gain Kp, it will accelerate the system response. Another approach is to use the deviation metric D as the objective function and employ numerical optimization algorithms such as gradient descent to solve online for the control parameters that reduce D, such as the adjustment direction and step size, and output the adjustment amounts ΔKp, ΔKi, and ΔKd.

[0140] Finally, the system assigns the calculated control parameter adjustments to the control algorithm in real time. Parameter updates follow the rule: Params new =Params old +α×ΔParams. Where Params old The set of parameters representing the internal parameters of the current control algorithm (such as Kp, Ki, Kd) is used. ΔParams is the set of adjustments output by the mapper. α is a learning rate factor (e.g., α=0.5) used to control the magnitude of each adjustment and prevent over-correction. After the parameters are updated, the system immediately adjusts based on the new parameter set Params. new The next step response test is executed automatically. The newly generated test data triggers a complete closed-loop process from data acquisition, model identification, feature comparison to parameter mapping. This iterative cycle continues until the comprehensive deviation metric D is less than the preset tolerance threshold, or the preset maximum number of iterations is reached. At this point, the control parameters have been automatically and adaptively tuned to a state that matches the dynamic characteristics of the component under test well, realizing rapid adaptive testing capabilities for components of different models and performance states. This process fundamentally replaces the traditional tuning mode that relies on manual trial and error based on engineer experience.

[0141] In some preferred embodiments, the method includes a manual testing mode, in which:

[0142] Provides a manually operated joystick connected to the mechanical transmission mechanism;

[0143] The use of the target motion trajectory command is stopped, and the real-time displacement of the joystick is used as the input of the main drive control component, while the generation and superposition of the torque compensation adjustment component are maintained.

[0144] During the manual operation of the joystick, the displacement data of the joystick and the actual torque data output by the electronic control load component are recorded simultaneously to form a displacement-torque relationship curve.

[0145] The displacement-torque relationship curve is compared and analyzed with a pre-stored standard curve conforming to aviation standards.

[0146] The method of the present invention further provides a manual testing mode, which aims to comprehensively evaluate the mechanical characteristics of the electronically controlled load components by combining subjective feelings with objective data through human-computer interaction.

[0147] When this mode is activated, the system performs the following technical steps. First, a manually operated joystick is provided, either physically directly coupled to the mechanical transmission mechanism or its own displacement sensor signal is designated as the system's primary input source. At this time, the control system software architecture switches modes: the use of preset target motion trajectory commands is completely discontinued at the software level, and the setpoint of the position control loop no longer originates from preset path points. Instead, the system uses the real-time displacement of the joystick as the input to the main drive control component. This means that the core task of the closed-loop position controller becomes driving the motor to follow the position changes manually applied to the joystick by the operator. Simultaneously, the generation and superposition logic of torque compensation adjustment components used to simulate specific damping, inertia, or spring effects remains effective. Therefore, the final comprehensive adjustment control signal applied to the servo motor is the dynamic sum of the position following command and the preset torque simulation command.

[0148] During manual operation of the joystick, including pushing, pulling, and rotating movements, the data acquisition system is simultaneously triggered. It records two key data streams in parallel at a high sampling rate: first, displacement data from position sensors on the joystick or mechanical transmission mechanism; second, actual torque data representing the output of the electronically controlled load assembly. This torque data can be directly measured by torque sensors installed in the drivetrain, or more generally, indirectly but accurately calculated by monitoring the phase current value output from the servo drive to the motor. These two synchronized data sequences are time-aligned and together form a dataset that can be plotted as a displacement-torque curve.

[0149] After testing, the measured displacement-torque curve is graphically overlaid and numerically compared with a pre-stored standard curve conforming to aviation standards within the system. For example, the standard curve might define a force feedback jump of a specific intensity at a particular displacement, or a linear increase in force induction with displacement throughout the entire stroke. The comparison analysis quantifies the deviation between the measured curve and the standard curve, such as maximum error and root mean square error, thereby providing a quantitative compliance assessment of the simulation fidelity of the force feedback system.

[0150] The method includes an automated mechanical performance testing mode, in which:

[0151] Receive input automated test parameters, which include at least the motion distance and the number of reciprocating motions;

[0152] Set the target torque curve command to zero or a constant value, and control the drive component to drive the component to reciprocate according to the automated test parameters;

[0153] During the reciprocating motion, the continuous output of the position sensor of the electronically controlled load assembly is collected and recorded to generate a position-time curve;

[0154] Analyze the smoothness and linearity of the position-time curve.

[0155] The method of this invention also includes an automated mechanical performance testing mode. This mode is designed to perform unattended, repeatable, quantitative testing of the fundamental physical properties of electronically controlled load components and related mechanical transmission systems, such as friction, clearance, and smoothness of motion.

[0156] In this mode, the testing process is as follows. First, the system receives automated test parameters input by the operator through its human-machine interface or configuration file. These parameters define at least the core boundary conditions of the test, such as the motion stroke, i.e., the starting and ending positions of the reciprocating motion; and the number of reciprocating motions, i.e., the number of test cycles. Other parameters may also include motion speed or acceleration.

[0157] Once the test begins, the control system sets the target torque curve command to zero or a constant reference value. Setting it to zero is typical for this mode, meaning that the torque compensation adjustment component is disabled, and the sole task of the drive component is to overcome the system's internal resistances (such as friction, cogging torque, etc.) to precisely execute the position command. Subsequently, the controller generates a trajectory command that reciprocates between the set motion strokes and controls the drive component to drive the component in reciprocating motion according to the automated test parameters.

[0158] Throughout the reciprocating motion, the data acquisition system collects and records the continuous outputs from the position sensors of the electronically controlled payload assembly. These position data points, collected at constant time intervals, constitute a time series, which can be plotted as a position-time curve.

[0159] Finally, the system performs data analysis on the acquired position-time curves, focusing on their smoothness and linearity. Linearity analysis is typically performed during the uniform motion phase, assessing the nonlinear characteristics of the system's friction by calculating the goodness of fit or deviation between the position-time curve and an ideal straight line. Smoothness analysis is more in-depth; it derives the velocity curve by taking the first derivative of the position-time curve and the acceleration curve by taking the second derivative. By analyzing the fluctuations and noise levels of the velocity or acceleration curves, or by performing spectral analysis, it can effectively identify minute vibrations caused by uneven gear meshing, bearing defects, or controller oscillations, thus providing a precise quantitative evaluation of the mechanical system's assembly quality and the smoothness of its dynamic response.

[0160] In this embodiment, the method further includes a comprehensive fault diagnosis step based on multi-source information cross-validation, specifically including:

[0161] The system simultaneously acquires position sensor signals and force sensor signals from the electronically controlled load assembly, as well as encoder signals from the drive component.

[0162] The acquired signals are compared and correlated with the driving commands issued to the driving components in real time.

[0163] Based on the predetermined matching rules and error tolerances that characterize the inherent physical correlation between signals when the system is working normally, determine whether the logical and numerical relationships between signals are within the normal range.

[0164] When it is determined that at least one logical relationship or numerical relationship exceeds the normal range, the system is deemed to have a functional fault, and a shutdown and alarm are triggered.

[0165] In this embodiment, to ensure absolute safety of the testing process and high reliability of the data, the method also integrates a comprehensive fault diagnosis step based on cross-validation of multi-source information. The core technology of this step lies in its transcendence of threshold monitoring of a single signal, and instead, by examining the dynamic correlation and physical consistency between multiple key signals within the system, it achieves real-time detection and response to deeper and more hidden functional faults.

[0166] The specific implementation method of this diagnostic step is as follows:

[0167] During each control cycle of the test system, the core control program performs synchronous data acquisition. This operation involves acquiring real-time data from three key, physically dispersed but logically strongly related sources. The first is the position sensor signal from the electronically controlled load assembly, which directly reflects the actual physical position of the system's final actuator. The second is the force sensor signal from the electronically controlled load assembly, which directly measures the actual physical torque output by the system. The third is the encoder signal from the drive component, which accurately reflects the angular position and motion state of the power source, namely the motor rotor.

[0168] These three key feedback signals, along with the system's forward command—the drive command issued by the main controller to the drive components—are compared and correlated in real time within a unified data processing framework. This drive command is typically a comprehensive adjustment and control signal that includes target position and compensation torque information. By using a high-precision, unified system clock to timestamp all signals, strict alignment of the data on the timeline is ensured, laying the foundation for subsequent comparison and judgment.

[0169] Subsequently, the system executes the core logic of correlation comparison. This logic is based on a series of predetermined matching rules and error tolerances, which are software-based representations of the inherent physical correlations between signals when the system is working normally.

[0170] One rule focuses on the consistency between commands and drives. It requires that the encoder feedback speed of the motor, that is, the first derivative of the position signal, should closely follow the motion components in the drive command within a very small time delay and tracking error range. If the command continuously issues motion commands but the encoder does not respond for a long time, it can be determined that the motor is locked or the drive is faulty.

[0171] Another rule focuses on the transmission consistency between the drive and the load. It requires that the displacement indicated by the encoder of the drive component, after being converted according to the transmission ratio, should maintain a strict correspondence with the position sensor signal of the electronically controlled load assembly. If the encoder signal shows that the motor is rotating while the position signal at the load end is stationary, it strongly indicates that mechanical decoupling has occurred, such as a broken belt or a slipped coupling.

[0172] Another rule focuses on the consistency between motion and mechanical response. It requires that, during system motion, the force sensor reading should match the sum of the torque compensation component in the drive command and the system's inherent resistance. If the command demands a significant damping force but the force sensor reading is close to zero, it may indicate a failure of the force sensor itself or its signal link.

[0173] In continuous real-time judgment, if the verification result of any of the above rules shows that the deviation of the logical or numerical relationship between two or more signals continuously or momentarily exceeds the preset error tolerance, for example, the displacement deviation between the drive end and the load end exceeds 0.5 mm, the system will trigger the fault judgment logic.

[0174] When at least one logical or numerical relationship is detected to be outside the normal range, the system immediately determines that a functional fault exists. This determination will immediately trigger a response procedure with safety as the highest priority. First, the system sends an emergency stop command to the drive components. A typical stop operation involves disabling the servo enable signal, i.e., executing a servo shutdown command, to instantly cut off the motor's torque output and activate any existing electromagnetic brakes to ensure the safe stopping of the mechanical system. Simultaneously, the system executes an alarm procedure, displaying a clear fault information window on the human-machine interface. The window will specifically explain the cause of the fault, such as indicating that a disconnection has been detected in the mechanical transmission chain between the drive end and the load end. Physical alarm indicator lights will also be illuminated, possibly accompanied by an audible warning. Critical multi-channel signal data before and after the fault, timestamps, and specific rule violations will be automatically recorded in a log file in non-volatile memory, providing accurate information for subsequent fault diagnosis and maintenance.

[0175] In this embodiment, before the method is executed, a step of parameter tuning of the control algorithm used to generate the main drive control component and the torque compensation adjustment component is included. This step is achieved by performing a series of step response tests.

[0176] Input the step-like target motion trajectory command into the system;

[0177] Record and analyze the time response curve of the actual motion position feedback, and extract its dynamic performance indicators such as overshoot, rise time and settling time;

[0178] Based on the dynamic performance index, the internal parameters of the control algorithm are iteratively adjusted until the dynamic performance index meets the preset performance requirements.

[0179] In this embodiment, before the method is formally executed, a crucial preparatory step of parameter tuning for the control algorithm is included. The purpose is to optimize and determine the optimal parameter configuration for the core control algorithm used to generate the main drive control component and the torque compensation adjustment component, thereby ensuring that the system exhibits excellent dynamic response characteristics and stability in subsequent testing tasks.

[0180] The parameter tuning steps are achieved by performing a series of standardized step response tests.

[0181] During testing, the system is first placed in a specific parameter tuning mode. In this mode, a step-like command for the target motion trajectory is input to the system. This command is characterized by its target position value abruptly changing from an initial value to an endpoint value in an instant, for example, jumping from 0mm to 50mm. This abrupt command can most effectively stimulate and expose the dynamic characteristics of the control system.

[0182] Under this step excitation, the system's data acquisition module records and analyzes the complete process of the actual motion position changing over time, as fed back by the position sensor, thus forming a time response curve. From this curve, a series of key dynamic performance indicators can be accurately extracted. These indicators mainly include overshoot, which reflects the system's stability and represents the maximum extent by which the actual position exceeds the target position; rise time, which measures the system's response speed and defines the time required for the position to rise from 10% to 90% of the target value; and settling time, which characterizes the system's convergence speed and refers to the time required for the position to enter and remain within a very small error band of the target value.

[0183] Based on these quantified dynamic performance metrics, technicians or an automated tuning program iteratively adjust the internal parameters of the control algorithm. These internal parameters are typically the proportional, integral, and derivative gains that form the core of the closed-loop control—the well-known PID parameters. After each parameter adjustment, the aforementioned step response test is repeated, and the dynamic performance metrics are re-evaluated. This iterative adjustment and testing cycle continues until the measured overshoot, rise time, and settling time fully meet a preset performance requirement. For example, the system might be required to have an overshoot of less than 5% and a settling time of less than 200ms when receiving a step command, in order to achieve a fast and smooth motion control effect.

[0184] See Figure 2 and Figure 3 The second embodiment of the present invention proposes an offline test platform for a flight simulator control load system, used for testing electronic control load components, the platform comprising:

[0185] The mechanical transmission mechanism includes a joystick 1, a drive motor 2, a first transmission component, a second transmission component, and a mechanical output end 3. The joystick 1 is drivenly connected to the input end of the first transmission component. The output end of the first transmission component is connected to the second transmission component and can drive the second transmission component as a whole to perform linear motion. The drive motor 2 is drivenly connected to the second transmission component. The mechanical output end 3 is disposed on the output end of the second transmission component and is used to connect to the output linkage of the electronically controlled load component.

[0186] The drive and control system includes a servo driver and a controller; the servo driver is electrically connected to the drive motor 2; the controller is connected to the position sensor and force sensor of the servo driver and the electronically controlled load assembly, respectively; the controller is used to receive feedback signals from the position sensor and force sensor, and generate drive commands according to test requirements, and control the action of the drive motor 2 through the servo driver.

[0187] This implementation establishes the basic framework of the offline testing platform. The mechanical transmission mechanism is the core physical component for realizing the testing actions. The joystick 1 serves as the direct interface for manual input, and its swing is used as the drive source for the first transmission component. The drive motor 2, as the only power source integrated into the platform, is the core actuator in the automatic testing mode.

[0188] The first and second transmission components are connected in series in a specific manner, enabling the first transmission component to drive the entire second transmission component module to produce linear displacement. The key to this design is that manual force can directly drive the entire second transmission component, including the drive motor 2, through the first transmission component. Simultaneously, the drive motor 2, integrated into the second transmission component, can independently drive the internal motion mechanism of that component.

[0189] Mechanical output terminal 3, as the final point of action, is reliably connected to the output link of the electronically controlled load assembly under test (ECU) via a connector such as a universal joint. The drive and control system is the central hub for the platform's intelligent and precise testing capabilities. Servo drives, such as the PacHP series industrial-grade products, are specifically designed to provide high-performance closed-loop drive for drive motor 2. The controller, typically composed of an industrial computer and a programmable logic controller (PLC), possesses multi-channel high-speed data acquisition and processing capabilities. It not only sends commands to the servo drive via a fieldbus, but more importantly, it establishes real-time signal connections with the position and force sensors inside the ECU via dedicated cables, thereby synchronously acquiring the real-time displacement and output torque of its output link. The host computer software running within the controller integrates a flight simulation model, capable of calculating the corresponding control load characteristics based on the input flight state parameters. Combining the position and torque signals fed back in real-time from the ECU, the controller performs real-time calculations using built-in advanced control algorithms to generate precise drive commands, which are then used by the servo drive to control the movement of drive motor 2.

[0190] The platform's workflow is as follows: In automatic testing mode, the controller generates target motion commands based on preset test scripts, and drives the drive motor 2 through a closed-loop control of the servo driver. The drive motor 2 drives the mechanical output end 3 through the second transmission component, thereby driving the output linkage of the component under test. During this process, the control system inside the component under test simulates the load according to the received commands (or its own model). The actual force and displacement generated are detected by its internal sensors and fed back to the platform controller, forming a closed loop for evaluating the performance of the component under test.

[0191] In manual testing mode, the tester directly operates joystick 1, whose movement directly drives the second transmission component and mechanical output end 3 via the first transmission component. The drive motor 2 can then follow the movement. The load simulation and data feedback process of the component under test is the same as in automatic mode. The advantage of this approach is that, through innovative mechanical series design, both manual and automatic input modes are organically integrated into a compact transmission mechanism. Simultaneously, a unified control system enables precise power control and comprehensive data acquisition, constructing a dedicated testing environment that is fully functional, structurally efficient, and precisely controlled. This provides a solid foundation for independent, efficient, and in-depth testing of electronically manipulated load components.

[0192] As a further explanation of the present invention, the first transmission assembly includes a first gear 4 and a first rack 5; the first gear 4 is connected to the control lever 1, the first rack 5 meshes with the first gear 4, and the first rack 5 is connected to the second transmission assembly.

[0193] This embodiment clarifies the preferred structure of the first transmission assembly. The first transmission assembly uses a rack and pinion mechanism to achieve efficient and reliable conversion from rotary motion to linear motion. Specifically, the control lever 1 is driven to the first gear 4 via a first universal joint 9. The function of the first universal joint 9 is to compensate for the angular deviation between the control lever and the input shaft of the first gear 4, ensuring smooth and unobstructed motion transmission. The reducer is used to increase the output torque, making manual operation easier and improving the precision of displacement control. The first rack 5 is precision-machined from high-strength alloy steel, maintaining precise meshing with the first gear 4, and its back is usually equipped with a slider, which cooperates with a high-precision linear guide fixed on the platform base to ensure the straightness of its motion trajectory and withstand lateral forces during operation. The end of the first rack 5 is rigidly fixed to the housing 6 of the second transmission assembly via a connecting plate. When the tester pushes or pulls the control lever 1, the swing is transmitted through the universal joint and converted into the rotation of the first gear 4, which in turn drives the meshing first rack 5 to make precise linear motion along the linear guide. The linear motion of the first rack 5 directly drives the housing 6 of the second transmission assembly to move as a whole through a fixed connection.

[0194] Its workflow is as follows: manual operation drives the lever 1 to swing, which is converted into the rotation of the first gear 4, then into the linear motion of the first rack 5, and finally into the linear translation of the entire second transmission assembly. The advantage of using a gear and rack mechanism is that it has high transmission rigidity, high efficiency, and a definite motion relationship. It can convert manual operation into large-stroke, high-precision linear drive without distortion or delay, providing a real, reliable, and highly repeatable physical input for manual function testing.

[0195] Furthermore, the second transmission assembly includes a housing 6, a second gear 7, and a second rack 8; the drive motor 2 is mounted on the housing 6, and its output shaft is connected to the second gear 7; the second rack 8 meshes with the second gear 7, and the second rack 8 is fixedly connected to the mechanical output end 3, which can move linearly relative to the housing 6; at least a portion of the housing 6 is connected to the first rack 5, so that the linear movement of the first rack 5 can drive the housing 6 and the second transmission assembly to move as a whole.

[0196] This embodiment details the internal structure of the second transmission assembly and its integration with the first transmission assembly and the drive motor 2. The second transmission assembly is a modular unit integrating power and actuation functions. The housing 6 is a robust frame, typically made of aluminum alloy or steel, serving as the load-bearing structure and mounting platform for the entire assembly.

[0197] The drive motor 2 is directly mounted on the outside of the housing 6 via its flange, and its output shaft is directly connected to the drive shaft of the second gear 7 via a rigid coupling. The second gear 7 and the second rack 8 constitute the second gear and rack pair inside the assembly. The second rack 8 is rigidly connected to the mechanical output end 3, which serves as the final output interface. The mechanical output end 3 is coupled to a precision linear guide rail mounted on the housing 6 via a linear bearing or linear slider, allowing the mechanical output end 3, together with the second rack 8, to perform high-precision, low-friction reciprocating linear sliding relative to the housing 6. The exterior of the housing 6 is provided with a dedicated connection interface, which is firmly connected to the end of the first rack 5, thus making the housing 6 and the first rack 5 a single unit in motion.

[0198] The linkage mechanism of this design manifests in two modes:

[0199] In purely manual mode, the movement of the first rack 5 drags the housing 6 to translate as a whole. At this time, the drive motor 2 can be de-energized or in position servo mode. The housing 6 drives the internal second gear 7, second rack 8, and mechanical output end 3 to move synchronously as a rigid body, with no relative movement between the mechanical output end 3 and the housing 6. In purely automatic mode, the drive motor 2 receives a controller command and rotates, driving the second gear 7 to rotate, thereby pushing the second rack 8 and its fixed mechanical output end 3 to move linearly relative to the stationary (or slowly moving) housing 6.

[0200] Its workflow has two paths: the manual drive path is "linear movement of the first rack 5 → overall translation of the housing 6 → synchronous translation of the mechanical output end 3"; the automatic drive path is "rotation of the drive motor 2 → rotation of the second gear 7 → linear movement of the second rack 8 relative to the housing 6 → linear movement of the mechanical output end 3". The core advantage of this implementation lies in creating a highly compact and functionally isolated dual-input drive module. It encapsulates all the elements required for automatic drive (motor, reduction gear, output rack) within a housing 6 that can be manually driven as a whole, achieving physical parallelism and functional decoupling of manual and automatic drive at the mechanical level. This design eliminates the need for complex clutch or switching devices, greatly simplifying the system structure, improving reliability, and ensuring that manual and automatic tests are applied through the same mechanical output end 3, guaranteeing the uniformity of test benchmarks and the comparability of test results.

[0201] In this embodiment, the platform further includes an installation mechanism, which includes a base plate and at least one clamp assembly; the base plate is provided with a positioning structure; the clamp assembly is detachably connected to the positioning structure and can be adjusted along the positioning structure to fix electronic control load components of different specifications.

[0202] This embodiment supplements the platform with a mounting mechanism for supporting and clamping the component under test, a key module for achieving the platform's versatility and convenience. The substrate is a heavy-duty metal platform with high rigidity and flatness, typically made of cast iron or pre-stretched aluminum alloy, serving as the mechanical reference surface for the entire testing equipment. Its surface is machined with a systematic positioning structure, preferably multiple T-slots running through the surface, or a matrix grid composed of densely arranged standardized threaded holes. The clamp assembly is a modular clamping tool, typically comprising a base with a T-nut, a vertically adjustable column, a horizontally rotatable and swingable pressure arm, and a quick-locking handle. The clamp assembly is inserted into the T-slot of the substrate via the T-nut at its bottom. After releasing the locking handle, it can slide freely along the T-slot direction; by lifting the T-nut and re-inserting it into an adjacent T-slot, it can move across slots, thus allowing arbitrary positioning in a two-dimensional plane. After positioning, tightening the handle securely fixes the clamp assembly to the substrate. In actual testing, based on the dimensions and fixing point location of the electronically controlled load assembly under test (such as ECL-110 or ECL-60), two or more clamp assemblies are moved to suitable positions around the assembly. The direction and height of the pressure arms are adjusted to ensure that the pressure blocks accurately act on the non-sensitive load-bearing surface of the assembly. Then, the handles are pressed down and locked to securely clamp the assembly onto the substrate. The workflow follows standardized procedures: determine the approximate clamping position based on the assembly model → place and pre-fix the clamp assemblies → suspend or place the assembly in the predetermined area of ​​the substrate → fine-tune the position of each clamp assembly to align the pressure heads → lock all clamps sequentially. The significant advantage of this implementation method is that it greatly expands the platform's versatility and operational efficiency. This fully adjustable modular installation scheme allows a single test platform to quickly adapt to and securely fix electronically controlled load assemblies of different models and sizes without the need for customized, expensive special tooling, significantly improving equipment utilization and economic efficiency. At the same time, it makes the installation, alignment and disassembly of the test piece very fast, simple and highly repeatable, greatly shortens the test preparation and changeover time, and ensures the geometric consistency of each installation, fundamentally reducing the test error caused by installation differences, and ensuring the reliability and validity of the test data.

[0203] The terms “first”, “second”, etc., are used to distinguish similar objects, not to describe or indicate a specific order or sequence.

[0204] The term "comprising" or any other similar term is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus / device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent in such process, method, article, or apparatus / device.

[0205] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. An offline testing method for a flight simulator control load system, characterized in that, The method includes the following steps: The electronically controlled payload assembly under test is mounted on the test platform; Based on the set simulated flight state parameters, generate target motion trajectory commands and target torque curve commands; Real-time acquisition of the actual motion position feedback and actual output torque feedback of the electronically controlled load assembly; The actual motion position feedback is compared with the target motion trajectory command, which is the main control target, to determine the position error, and the main drive control component is generated based on the position error. The actual output torque feedback is compared with the target torque curve command, which serves as an auxiliary correction target, to determine the torque error, and a torque compensation adjustment component is generated based on the torque error. The torque compensation adjustment component is superimposed on the main drive control component to form a comprehensive adjustment control signal; The test platform's drive components analyze and execute the integrated adjustment and control signals to drive the mechanical transmission mechanism connected to the electronically controlled load assembly.

2. The offline testing method for a flight simulator control load system according to claim 1, characterized in that, The method further includes a transmission delay measurement step, specifically comprising: In response to the input step motion command, record the first time point at which the step motion command was issued; Continuously monitor the actual motion position feedback; When the value of the actual movement position feedback is detected to reach the preset trigger threshold for the first time, the corresponding second time point is recorded; The motion response delay time of the electronically controlled load assembly is obtained by calculating the difference between the second time point and the first time point.

3. The offline testing method for a flight simulator control load system according to claim 1, characterized in that, The method includes a manual testing mode, in which: Provides a manually operated joystick connected to the mechanical transmission mechanism; The use of the target motion trajectory command is stopped, and the real-time displacement of the joystick is used as the input of the main drive control component, while the generation and superposition of the torque compensation adjustment component are maintained. During the manual operation of the joystick, the displacement data of the joystick and the actual torque data output by the electronic control load component are recorded simultaneously to form a displacement-torque relationship curve. The displacement-torque relationship curve is compared and analyzed with a pre-stored standard curve conforming to aviation standards.

4. The offline testing method for a flight simulator control load system according to claim 1, characterized in that, The method includes an automated mechanical performance testing mode, in which: Receive input automated test parameters, which include at least the range of motion and the number of reciprocating motions; Set the target torque curve command to zero or a constant value, and control the drive component to drive the electronically controlled load assembly to reciprocate according to the automated test parameters; During the reciprocating motion, the continuous output of the position sensor of the electronically controlled load assembly is collected and recorded to generate a position-time curve; Analyze the smoothness and linearity of the position-time curve.

5. The offline testing method for a flight simulator control load system according to claim 1, characterized in that, The method also includes a comprehensive fault diagnosis step based on multi-source information cross-validation, specifically including: The system simultaneously acquires position sensor signals and force sensor signals from the electronically controlled load assembly, as well as encoder signals from the drive component. The acquired signals are compared and correlated with the driving commands issued to the driving components in real time. Based on the predetermined matching rules and error tolerances that characterize the inherent physical correlation between signals when the system is working normally, determine whether the logical and numerical relationships between signals are within the normal range. When it is determined that at least one logical relationship or numerical relationship exceeds the normal range, the system is deemed to have a functional fault, and a shutdown and alarm are triggered.

6. The offline testing method for a flight simulator control load system according to claim 1, characterized in that, Before the method is executed, a step of parameter tuning of the control algorithm used to generate the main drive control component and the torque compensation adjustment component is included. This step is achieved by performing a series of step response tests. Input the step-like target motion trajectory command into the system; Record and analyze the time response curve of the actual motion position feedback, and extract its dynamic performance indicators such as overshoot, rise time and settling time; Based on the dynamic performance index, the internal parameters of the control algorithm are iteratively adjusted until the dynamic performance index meets the preset performance requirements.

7. An offline testing platform for implementing the offline testing method for a flight simulator control load system as described in any one of claims 1-6, characterized in that, The offline testing platform includes: The mechanical transmission mechanism includes a control lever, a drive motor, a first transmission component, a second transmission component, and a mechanical output end; The joystick is driven to the input end of the first transmission component; The output end of the first transmission component is connected to the second transmission component and can drive the second transmission component as a whole to perform linear motion; The drive motor is connected to the second transmission component in a driving connection; The mechanical output end is located on the output end of the second transmission component and is used to connect the output link of the electronically controlled load component. Drive and control systems, including servo drives and controllers; The servo driver is electrically connected to the drive motor; The controller is connected to the position sensor and force sensor of the servo driver and the electronically controlled load assembly, respectively. The controller is used to receive feedback signals from the position sensor and force sensor, and generate drive commands according to test requirements, and control the action of the drive motor through the servo driver.

8. The offline testing platform according to claim 7, characterized in that, The first transmission assembly includes a first gear and a first rack; The first gear is connected to the control lever, the first rack meshes with the first gear, and the first rack is connected to the second transmission assembly.

9. The offline testing platform according to claim 8, characterized in that, The second transmission assembly includes a housing, a second gear, and a second rack; The drive motor is mounted on the housing, and its output shaft is connected to the second gear transmission. The second rack meshes with the second gear, and the second rack is fixedly connected to the mechanical output end, which can move linearly relative to the housing; At least a portion of the housing is connected to the first rack, such that the linear motion of the first rack can drive the housing and the second transmission assembly to move as a whole.

10. The offline testing platform according to claim 7, characterized in that, The platform also includes a mounting mechanism, which includes a base plate and at least one clamp assembly; The substrate is provided with a positioning structure; The clamp assembly is detachably connected to the positioning structure and can be adjusted along the positioning structure to fix electronically controlled load components of different specifications.

Images