A simulation method for bench testing of a continuously variable transmission

Abstract

The application discloses a simulation method for bench test of a continuously variable transmission, comprising the following steps: constructing a parameterized scene model containing a driving scene; calculating the required power of a transmission system under the driving scene through forward simulation based on the parameterized scene model; adopting a global optimization algorithm, wherein the theoretical optimal control sequence at least comprises a theoretical optimal continuously variable transmission speed ratio sequence; reproducing the driving scene on a test bench; and calculating a preset multi-dimensional performance difference index based on the aligned data. Thus, an analysis method is provided, which generates a theoretical benchmark based on forward simulation and global optimization, and dynamically compares the benchmark with high-precision bench test data in multiple dimensions. The software and hardware are universal and structured, and can be more widely adapted to continuously variable transmission systems of different sizes, functions and loads, while simulating the operation conditions of different scenes, so as to realize the integration of scene simulation and bench test.

Description

Technical Field

[0001] This invention relates to the field of transmission simulation testing technology, and in particular to a simulation method for continuously variable transmission (CVT) bench testing. Background Technology

[0002] Continuously variable transmission (CVT) systems are a very important field in mechanical transmission. However, traditional testing platforms are often customized for a specific CVT system and lack a general-purpose structured testing platform supported by scenario simulation systems.

[0003] Currently, when using dedicated continuously variable transmission (CVT) system test platforms, various CVT systems typically have different characteristic dimensions and power characteristics, usually requiring customized test benches and corresponding software. Setting up general-purpose simulation software requires configuring extremely complex physical simulation engines, while dedicated simulation systems need to be custom-developed or self-developed, resulting in relatively high costs. Furthermore, multiphysics modeling and operation are necessary to achieve structural optimization in a virtual environment, and then the completed CVT system prototype must undergo dedicated bench testing.

[0004] Therefore, there is a need for a method that can accurately test continuously variable transmission (CVT) systems without significantly increasing software complexity. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a simulation method for bench testing of continuously variable transmissions (CVTs).

[0006] The objective of this invention is achieved through the following technical solution:

[0007] This application discloses a simulation method for bench testing of continuously variable transmissions (CVTs), comprising the following steps: constructing a parameterized scenario model containing a driving scenario; calculating the required power of the transmission system under the driving scenario through forward simulation based on the parameterized scenario model; employing a global optimization algorithm, with the goal of minimizing the total system loss, solving for the theoretically optimal control sequence under the driving scenario based on the required power, wherein the theoretically optimal control sequence includes at least the theoretically optimal CVT speed ratio sequence; reproducing the driving scenario on a test bench, driving a physical prototype of the CVT, and synchronously collecting measured data of the physical prototype during operation, wherein the measured data includes at least the torque and speed of the input and output shafts; synchronizing the measured data with the theoretically optimal control sequence in time, and calculating a preset multidimensional performance difference index based on the aligned data.

[0008] Its beneficial effects are:

[0009] This paper presents an analytical method based on forward simulation and global optimization to generate theoretical benchmarks, which are then dynamically compared with high-precision bench test data in multiple dimensions. The software and hardware are universal and structured, making it more widely adaptable to continuously variable transmission (CVT) systems of different sizes, functions, and loads. It also simulates different operating conditions, achieving integrated scenario simulation and bench testing. Furthermore, it enables the reuse and portability of data assets, providing strong support for subsequent AI intelligent analysis and AI training datasets. This allows for closed-loop evaluation and precise root cause analysis of CVT performance, significantly improving R&D efficiency and optimization focus. By establishing a theoretically optimal benchmark, it provides an objective standard for evaluating CVT control strategies. In addition, through multi-dimensional error decomposition, the total efficiency loss can be accurately located to mechanical, hydraulic, or control subsystems, guiding targeted optimization and achieving a closed loop between digital simulation and physical testing, thus shortening the R&D iteration cycle.

[0010] Furthermore, the multidimensional performance difference indicators include: the tracking error of the measured speed ratio relative to the theoretical optimal speed ratio command; the instantaneous difference and cumulative difference between the measured transmission efficiency and the theoretical optimal transmission efficiency; and the actual response time of speed ratio changes.

[0011] Furthermore, solving for the theoretically optimal control sequence includes the following sub-steps: discretizing the optimization problem to construct a state space containing vehicle speed and continuously variable transmission (CVT) ratio, and a decision space containing engine torque and CVT ratio change rate; defining the state transition equation from the current time step to the next time step; defining a single-step cost function containing engine fuel consumption rate and smoothing penalty term; iterating backward from the scenario endpoint to solve for the minimum cumulative cost and corresponding optimal decision for each state; and backtracking forward from the initial state to obtain the globally optimal engine torque sequence, engine speed sequence, and CVT ratio sequence.

[0012] Furthermore, after calculating the multidimensional performance difference index, the method further includes: calculating the total power loss of the continuously variable transmission based on the measured data or simulation data; decomposing the total power loss into multiple sub-losses, including at least slip loss, oil churning loss, bearing friction loss, and hydraulic pump loss; comparing the proportion or cumulative energy of each sub-loss in the total loss to perform a loss contribution analysis in order to locate performance bottlenecks.

[0013] Furthermore, the test bench is equipped with a universal hardware interface, which is used to adapt to continuously variable transmissions of different sizes and models.

[0014] Furthermore, the data defined by the parameterized scenario model includes a demand speed sequence and an environmental load sequence; wherein the environmental load sequence includes changes in road slope and / or wind speed over time.

[0015] Furthermore, the forward simulation calculation of the required power of the transmission system in the driving scenario includes: calculating the total resistance of the vehicle at each moment based on the scenario data, wherein the total resistance includes acceleration resistance, slope resistance, rolling resistance and air resistance; calculating the required wheel-side power based on the total resistance and the required vehicle speed at the current moment; and combining the transmission chain efficiency to back-engineer the required wheel-side power to the input end of the continuously variable transmission to obtain the theoretical required power sequence of the engine.

[0016] Furthermore, the theoretically optimal control sequence and the measured data are stored in a database respectively, and associated with each other through scene identifiers and time series to establish a traceable correspondence between simulation data and measured data. Attached Figure Description

[0017] Figure 1 This is a simplified flowchart illustrating the simulation method for continuously variable transmission (CVT) bench testing according to some embodiments of this application.

[0018] Figure 2 This is a simplified schematic diagram of a simulation method for continuously variable transmission (CVT) bench testing according to some embodiments of this application;

[0019] Figure 3 The following is a simplified flowchart of the test bench control logic for some embodiments of this application. Detailed Implementation

[0020] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] A simulation method for bench testing of continuously variable transmissions (CVTs) according to an embodiment of this application aims to solve the problems in the prior art where the difference between theoretical optimal performance and actual performance cannot be effectively separated and quantified during CVT platform testing, resulting in unclear optimization direction, lack of comprehensive evaluation of dynamic cyclic scenarios, isolated simulation and testing, and single performance evaluation index, making it difficult to pinpoint the specific source of efficiency loss.

[0022] In this embodiment, reference Figure 1 and Figure 2 The simulation method for continuously variable transmission (CVT) bench testing includes the following steps:

[0023] S1. Construct a parameterized scenario model that includes driving scenarios for simulation task customization. This mainly includes setting application scenarios and CVT system parameters, power and load parameters.

[0024] S2. Based on the parameterized scenario model, calculate the required power of the transmission system under the driving scenario through forward simulation.

[0025] S3. Employing a global optimization algorithm, with the goal of minimizing total system loss, the theoretically optimal control sequence for the driving scenario is solved based on the required power. This theoretically optimal control sequence includes at least the theoretically optimal continuously variable transmission (CVT) speed ratio sequence, and theoretically optimal parameter settings are established.

[0026] S4. Reproduce the driving scenario on a test bench, drive the continuously variable transmission (CVT) physical prototype, and simultaneously collect measured data of the physical prototype during operation. The measured data includes at least the torque and speed of the input and output shafts. Figure 1 The CVT system uses a customized interface, i.e., a universal hardware interface, to connect to the CVT continuously variable transmission. In this example, multiple solutions and customized parts can be selected based on technical options, followed by installation and integration. The data is converted from digital to analog based on the S3 test bench control signals and input to the test bench for simulation testing. Then, the sensor data is converted from analog to digital and sent to the data storage unit. The data is then processed through a data preprocessing unit for test result analysis. Furthermore, each round of simulation testing is evaluated; if the current round of testing is successful, the next round of hardware adaptation begins. Test result analysis includes the presentation of measured data and a result analysis unit, a simulation test report and technical selection parameter optimization suggestions, test data cleaning, labeling, and grouping.

[0027] S5. The measured data and the theoretically optimal control sequence are synchronized and aligned in time, and a preset multi-dimensional performance difference index is calculated based on the aligned data. Through the above steps, the method of this application realizes a refined comparison and analysis between the theoretical optimal and the measured performance, thereby providing a complete data closed loop for the performance testing and precise optimization of the continuously variable transmission (CVT) system.

[0028] In this embodiment, a parameterized model of driving scenarios, including vehicles and drones, is constructed. This parameterized scenario model defines the boundaries and input conditions of the simulation. Specifically, the scenario definition data includes: scenario identifiers, used to uniquely identify different test scenarios; time series, defining the time axis of the simulation; demand speed series, describing the change of the target speed of the tested system (e.g., vehicle) over time in this scenario, representing driving intentions or task requirements; environmental load series, describing the external conditions affecting system operation, such as changes in road gradient and wind speed over time; and system load parameters, defining the inherent load characteristics applied to the tested system in the scenario, such as vehicle mass, road rolling resistance coefficient, and air resistance coefficient. In addition, simulation modeling requires a series of data definitions, including: power source characteristic data, such as the universal characteristic map of the engine, to describe its torque and fuel consumption characteristics at different speeds and throttle openings; transmission characteristic data, including the working speed ratio range of the continuously variable transmission (CVT) and its efficiency characteristic map, which reflects the transmission efficiency of the transmission at different input speeds, input torques and speed ratios; and vehicle or platform transmission parameters, such as the final drive ratio and wheel radius, to translate the vehicle requirements into the requirements of the transmission system.

[0029] Subsequent steps are performed based on the theoretically optimal control sequence and theoretical performance indicators. The theoretically optimal control sequence is, under the given scenario, the theoretically optimal speed ratio sequence, the optimal engine torque sequence, and the speed sequence calculated using an optimization algorithm (such as dynamic programming) to minimize the total system loss or total energy consumption. The theoretical performance indicators are, based on the theoretically optimal control sequence, the theoretical total power loss, theoretical fuel consumption rate, etc.

[0030] To better understand this, the simulation method for continuously variable transmission (CVT) bench testing will be explained in detail below.

[0031] In step S2, based on the scenario model and vehicle dynamics model, the required power is calculated through forward simulation. Specifically, according to a preset driving scenario, the vehicle's dynamic behavior is simulated, and the required power of the transmission system is calculated time-by-time. The execution steps are as follows:

[0032] Step S21: Calculate the total resistance of the vehicle.

[0033] At each simulation time step Calculate the total resistance currently experienced by the vehicle based on scenario data. The unit is Newton (N). The calculation formula is as follows:

[0034]

[0035] in, To increase resistance. The vehicle's equivalent mass (unit: kg) includes the inertial mass of rotating components; The demand acceleration at the current moment (unit: m / s²) is obtained by differentiating the demand velocity sequence in the scenario definition data. This refers to the slope resistance. Acceleration due to gravity (unit: m / s²); The current road gradient (in %) is derived from the scene definition data. This represents rolling resistance. The rolling resistance coefficient (dimensionless) is one of the system load parameters. This refers to air resistance. Air density (unit: kg / m³); is the drag coefficient (dimensionless). The vehicle's frontal area (unit: m²) is a system load parameter. The required vehicle speed at the current moment (unit: m / s); The wind speed at the current moment (unit: m / s) is derived from the scene definition data.

[0036] Step S22: Calculate the required wheel-side power;

[0037] Calculate the power required by the transmission system at the wheel ends based on the total resistance and vehicle speed. The unit is kilowatt (kW). The calculation formula is as follows:

[0038]

[0039] Step S23: Reverse the theoretical requirements to the input of the transmission:

[0040] By combining the wheel-side power demand with the drivetrain efficiency, we can reverse-engineer the input of the continuously variable transmission (CVT). This allows us to calculate the theoretical power requirement that the engine needs to provide at the current moment. The unit is kilowatt (kW). The simplified calculation formula is as follows:

[0041]

[0042] in, The efficiency of the main reducer (dimensionless) is part of the system and component parameter data. This represents the overall efficiency (dimensionless) of other transmission components (such as the differential). The algorithm's output is a time series. It serves as the input to the optimization algorithm in subsequent steps, representing the basic power flow necessary to meet the driving requirements of the scenario, without considering the specific engine and CVT operating point.

[0043] In step S3, a global optimization algorithm is used to solve for the theoretically optimal engine operating point sequence and CVT speed ratio sequence with the goal of minimizing the total system loss. The global optimization algorithm in step S3 is either a dynamic programming algorithm or the Pontryagin minimum principle.

[0044] Specifically, with the goal of minimizing the total system energy consumption or total loss, the theoretically optimal control command sequence is calculated for a known complete driving scenario. This invention preferably utilizes a dynamic programming algorithm, with the following steps:

[0045] Step S31: Construct the state space and decision space;

[0046] Discretize the optimization problem and define the state variable as vehicle speed. (Unit: m / s) and continuously variable transmission (CVT) speed ratio (Dimensionless), define the control variable as engine torque. (Unit: Nm) and speed ratio change rate (Unit: s⁻¹).

[0047] Step S32: Define the state transition equation;

[0048] Establish from time arrive The state transition relationship is as follows. The vehicle speed transition is determined by the vehicle dynamics model and is related to engine torque, speed ratio, vehicle resistance, etc. The speed ratio transition is approximately described by the following formula:

[0049]

[0050] in, This represents the discrete time step (unit: s).

[0051] Step S33: Define the single-step cost function;

[0052] Calculate from state Departure, decision-making The cost of transitioning to the next state The unit is grams of fuel or joules of energy. The calculation formula is as follows:

[0053]

[0054] in, This is the engine fuel consumption rate function (unit: g / s), obtained by looking up the engine universal characteristic diagram. Engine speed. (Unit: rpm) and It is calculated based on the transmission relationship. This is a smoothing penalty used to avoid drastic changes in the speed ratio. The penalty weighting coefficient (unit: g·s or J·s).

[0055] Step S34: Solve by reverse iteration;

[0056] From the end moment of the scene Begin by iterating backwards to the starting point. For each discrete state point, calculate and store the minimum cumulative cost to the destination. and the corresponding optimal decision The recursive formula is:

[0057]

[0058] Step S35: Backtrack forward to obtain the optimal trajectory;

[0059] Starting from the initial state, based on the optimal decision stored at each step... By backtracking forward, the globally optimal control sequence can be obtained: the theoretically optimal engine torque sequence. Theoretical optimal engine speed sequence And the theoretically optimal CVT speed ratio sequence .

[0060] After obtaining the theoretically optimal control sequence, this method reproduces the driving scenario on a test bench and synchronously collects measured data of speed, torque, and power at both the input and output ends. The measured data is acquired in real-time by sensors on a general-purpose test bench and includes at least: a high-precision synchronous timescale to provide a unified time reference for all collected data; dynamic signals from the input and output shafts, including real-time torque and speed signals from the continuously variable transmission (CVT) input and output shafts; and transmission operating status signals, including real-time signals directly related to the transmission's operating status, such as hydraulic clamping pressure and lubricating oil temperature. Figure 3 As shown, the specific process for acquiring the measured data includes: the simulation model in the host computer calculates the torque command T that should be applied to the CVT at the current moment. cmd Speed ​​command N cmd and speed ratio instruction R cmd Then, the real-time controller executes control logic to simulate engine output, executes control logic to change the speed ratio by adjusting the pulley displacement, and executes control logic to calculate and apply load torque in real time.

[0061] In detail, the torque command T applied to the CVT cmd Speed ​​command N cmd and speed ratio instruction R cmdAs digital signals, the output is standard industrial analog signals via the D / A conversion module of the real-time controller. For example, ±10V voltage or 4-20mA current corresponds to torque / speed commands sent to the motor servo driver, while 0-10V voltage corresponds to proportional valve pressure commands controlling the CVT clamping force. The physical meaning, such as 1V potentially corresponding to 100 Nm of torque or 500 rpm of speed, is calibrated by the driver parameters. Regarding sensor data acquisition, torque sensors typically output two sine waves (mV level) with a 90° phase difference or internally calculated ±5V / ±10V analog voltages, with the voltage value linearly related to the torque. Speed ​​encoders output TTL pulse sequences (digital signals), with frequency proportional to the speed. Temperature / pressure sensors typically output 4-20mA current signals. These signals are conditioned; for example, torque signals are amplified, filtered, and anti-aliasing processed by a dynamic strain gauge, while current signals are converted to voltage signals via precision sampling resistors. All conditioned analog voltage signals are then connected to a multi-channel synchronous acquisition card for synchronous A / D acquisition. Key parameters include sampling rate (typically > 1kHz), resolution (16-bit or 24-bit), and synchronization accuracy (all channels use the same sampling clock with an error <1μs). A hardware trigger signal to start the test is issued by the bench controller, ensuring that all acquisition devices have a consistent time base. Finally, the acquisition card transmits the A / D converted digital array to an industrial computer in real time via PCIe or Ethernet. The industrial computer runs real-time data logging software (such as NI DIAdem, ETAS INCA), adds a unified timestamp to the data, and writes it to a high-speed solid-state drive, often in binary file format (such as TDMS). After the test, a post-processing service parses the binary file, downsamples it (e.g., from 1kHz to 100Hz), and stores it in the corresponding measured data table in a central relational database (such as MySQL / PostgreSQL), linking it to the simulation theoretical data table via ScenarioID and Time.

[0062] In S4, the scenario is reproduced on the test bench, and the measured data of speed, torque and power at the power input and output ends are collected synchronously.

[0063] For example, physical prototype operating data is acquired in real time from sensors on a general-purpose test bench. This includes at least: a high-precision synchronization timescale, providing a unified time reference for all acquired data; dynamic signals from the input and output shafts, including real-time acquired torque and speed signals from the continuously variable transmission (CVT) input and output shafts; and transmission operating status signals, including real-time acquired signals directly related to the transmission's operating status, such as hydraulic clamping pressure and lubricating oil temperature.

[0064] In step S5, the measured data and the theoretically optimal sequence are synchronized in time, and a preset multidimensional performance difference index is calculated. For example, the analysis results generated after synchronizing and calculating the theoretical benchmark data and the measured data are used to quantify the performance gap. The multidimensional performance difference index includes at least: control tracking error, such as the tracking error of the measured speed ratio relative to the theoretically optimal speed ratio command; energy efficiency difference index, such as the instantaneous and cumulative difference between the measured transmission efficiency and the theoretically optimal transmission efficiency; and dynamic response index, such as the actual response time of speed ratio changes. These indexes allow for the precise quantification of the performance gap between the physical prototype and the theoretical optimum.

[0065] For example, the analysis results generated after time synchronization and calculation of the theoretical benchmark data and the measured data are used to quantify performance differences. These include at least: control tracking error, such as the tracking error of the measured speed ratio relative to the theoretical optimal speed ratio command; energy efficiency difference indicators, such as the instantaneous and cumulative differences between the measured transmission efficiency and the theoretical optimal transmission efficiency; and dynamic response indicators, such as the actual response time of speed ratio changes.

[0066] The definition, generation, collection, and comparative analysis of the above five types of data constitute a complete data loop from virtual simulation to physical measurement and then to quantitative evaluation, supporting the performance testing and precise optimization of continuously variable transmission (CVT) systems of different specifications on a generalized platform.

[0067] Therefore, the total power loss obtained from actual measurement or simulation can be decomposed into different physical sources, enabling accurate root cause analysis. The calculation is based on a series of sub-models, and the steps are as follows:

[0068] Step S51: Calculate the total transmission efficiency and total power loss;

[0069] Calculate the total transmission efficiency of the continuously variable transmission (CVT) based on measured or simulated input / output data. (Dimensionless) and total power loss (Unit: W)

[0070]

[0071] in, Input / output shaft torque (unit: Nm). Angular velocity of the input / output shaft (unit: rad / s).

[0072] Step S52: Calculate the individual losses;

[0073] The total power loss is modeled as the sum of several main component losses, and each is calculated separately:

[0074]

[0075] In the above formula, slip loss This is caused by microscopic slippage between the transmission belt (or chain) and the pulley. Its model is as follows:

[0076]

[0077] in, The minimum coefficient of friction (dimensionless) required to transmit the current torque, and the clamping force (Unit: N) Related; The actual coefficient of friction (dimensionless). Belt speed (unit: m / s).

[0078] Oil churning loss It is generated by the rotating component agitating the lubricating oil. Its model is:

[0079]

[0080] in, The oil stirring loss coefficient (dimensionless); The density of the lubricating oil (unit: kg / m³); Characteristic diameter (unit: m); To match oil temperature (Unit: °C) Related viscosity correction function.

[0081] Bearing friction loss This is generated by the rotation of the bearing under stress. Its simplified model is:

[0082]

[0083] in, For the relevant shaft speed (unit: rad / s); The bearing friction coefficient is dimensionless.

[0084] Hydraulic pump loss It provides the hydraulic power required for clamping force and lubrication. Its model is as follows:

[0085]

[0086] in, The required flow rate of the pump (unit: m³ / s); The system pressure (unit: Pa) is usually determined by the measured clamping pressure. reflect; The efficiency of the pump (dimensionless).

[0087] Step S53: Loss Contribution Analysis;

[0088] By comparing the losses of each item In total losses The proportion of energy loss, or the cumulative energy loss obtained by integrating it over the entire test scenario. This allows for quantitative analysis of the contribution of each component to the overall efficiency, thereby pinpointing performance bottlenecks.

[0089] In other words, in this embodiment, the method is executed based on the software interface of the simulation platform and the hardware interface of the test platform. The software interface is used for the customization of the simulation environment and control strategy for the application scenario, as well as the data analog / analog-to-digital signal conversion, identification, and presentation and analysis of test results. The hardware interface of the test platform provides a universal hardware interface conversion for continuously variable transmission (CVT) systems of different sizes, and maximizes compatibility with various models of CVT systems, thereby greatly reducing the cost of dedicated test benches.

[0090] The main application scenarios include modeling the dynamic characteristics of any continuously variable transmission (CVT) system in use. Taking automobiles as an example, there are urban operating conditions, highway operating conditions, continuous uphill operating conditions, and mountain operating conditions. To simulate these different operating conditions, it is necessary to simulate the operation, sensing, and control of the CVT system. At the same time, theoretical optimal values ​​corresponding to these operating conditions can be set for subsequent comparative analysis and targeted improvement. Taking drones as an example, CVT system operating condition simulation is also required in different scenarios such as takeoff, landing, and crosswind disturbances. After such simulation is converted into control signals, it performs transmission and control operations on the real CVT system and collects measured data for analysis and comparison. This provides data support for CVT system technology selection, product development and improvement, fault diagnosis, and bottleneck optimization. Its advantages are that it greatly shortens the development cycle and significantly reduces development and testing costs. Due to the reusable and portable nature of simulation scenario settings, test environments, results, and other data assets, knowledge can be transferred to further reduce costs and increase efficiency.

[0091] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A simulation method for bench testing of continuously variable transmissions (CVTs), characterized in that, Includes the following steps: Construct a parameterized scene model that includes driving scenarios; Based on the parameterized scenario model, the power demand of the transmission system under the driving scenario is calculated through forward simulation. A global optimization algorithm is used to minimize the total system loss. The theoretically optimal control sequence for the driving scenario is solved based on the required power. The theoretically optimal control sequence includes at least the theoretically optimal continuously variable transmission (CVT) speed ratio sequence. The driving scenario was reproduced on a test bench, and a physical prototype of the continuously variable transmission was driven to run. Simultaneously, measured data of the physical prototype during operation were collected. The measured data included at least the torque and speed of the input shaft and the output shaft. The measured data is time-synchronized and aligned with the theoretically optimal control sequence, and a preset multidimensional performance difference index is calculated based on the aligned data.

2. The simulation method for continuously variable transmission (CVT) bench testing according to claim 1, characterized in that, The multidimensional performance difference indicators include: The tracking error of the measured speed ratio relative to the theoretical optimal speed ratio command; Instantaneous and cumulative differences between measured transmission efficiency and theoretical optimal transmission efficiency; The actual response time of speed ratio change.

3. The simulation method for continuously variable transmission (CVT) bench testing according to claim 1 or 2, characterized in that, Solving for the theoretically optimal control sequence involves the following sub-steps: The optimization problem is discretized to construct a state space containing vehicle speed and continuously variable transmission (CVT) speed ratio, and a decision space containing engine torque and speed ratio change rate. Define the state transition equation from the current time step to the next time step; Define a single-step cost function that includes engine fuel consumption rate and smoothing penalty term; Iterate backward from the end of the scenario to find the minimum cumulative cost and the corresponding optimal decision for each state; By backtracking from the initial state, we obtain the globally optimal engine torque sequence, engine speed sequence, and continuously variable transmission (CVT) speed ratio sequence.

4. The simulation method for continuously variable transmission (CVT) bench testing according to claim 2, characterized in that, After calculating the multidimensional performance difference index, the following is also included: Calculate the total power loss of the continuously variable transmission based on the measured or simulated data. The total power loss is decomposed into at least several sub-losses, including slip loss, churning loss, bearing friction loss, and hydraulic pump loss. By comparing the proportion or cumulative energy of each sub-item loss in the total loss, a loss contribution analysis is conducted to pinpoint performance bottlenecks.

5. The simulation method for continuously variable transmission (CVT) bench testing according to claim 1, characterized in that, The test bench is equipped with a universal hardware interface, which is used to adapt to continuously variable transmissions of different sizes and models.

6. The simulation method for continuously variable transmission (CVT) bench testing according to claim 1, characterized in that: The data defined by the parameterized scenario model includes a demand speed sequence and an environmental load sequence; wherein, the environmental load sequence includes the changes in road slope and / or wind speed over time.

7. The simulation method for continuously variable transmission (CVT) bench testing according to claim 6, characterized in that, The calculation of the required power of the transmission system under the driving scenario through forward simulation includes: The total resistance of the vehicle is calculated time-by-time based on the scene data. The total resistance includes acceleration resistance, slope resistance, rolling resistance and air resistance. Calculate the required wheel-side power based on the total resistance and the required vehicle speed at the current moment; By combining the transmission chain efficiency, the required wheel-side power is reversed to the input end of the continuously variable transmission (CVT) to obtain the theoretical required power sequence of the engine.

8. The simulation method for continuously variable transmission (CVT) bench testing according to claim 1, characterized in that, The theoretically optimal control sequence and the measured data are stored in a database respectively, and associated with each other through scene identifiers and time series to establish a traceable correspondence between simulation data and measured data.

Images