A mechanical inertia compensation method for high-precision dynamic simulation-oriented electric vehicle test bench

Abstract

The application discloses a kind of high-precision dynamic simulation-oriented mechanical inertia compensation method of electric vehicle test bench.This method introduces extended state observer in inertia compensator to replace traditional differential element, the total disturbance of system including acceleration dynamics and unmodeled dynamics is expanded to new state variable to carry out unified observation and compensation, solve the problem that high frequency noise amplification of test bench under high multiple compensation leads to the decline of speed simulation accuracy;The system closed-loop discrete mathematical model containing ESO is constructed, and the stability boundary of the system under high multiple compensation is determined using its characteristic equation and Jury criterion;In the stability domain, the sparrow search algorithm with working condition self-adaptive boundary early warning mechanism is used to optimize the observer gain, and the conflict between low-speed high-precision requirement and high-speed robustness requirement is solved.The application effectively solves the interference of high-frequency noise on the compensation loop, and realizes the high-precision and high-stability simulation of the mechanical dynamics of large inertia electric vehicle under all working conditions.

Description

Technical Field

[0001] This invention belongs to the field of electric vehicle testing technology, and in particular, it is a method for compensating the mechanical inertia of an electric vehicle test bench for high-precision dynamic simulation. Background Technology

[0002] With the continuous advancement of electric vehicle technology and the development of market demand, electric vehicles are facing increasingly stringent requirements regarding their operational performance, such as safety, reliability, energy management efficiency, and powertrain response. This necessitates thorough testing and verification during the vehicle development phase. However, conducting on-site testing of electric vehicles in real-world road environments presents challenges, including high testing costs, uncontrollable environmental conditions, significant safety risks, long testing cycles, and difficulty in accurately reproducing operating conditions. Therefore, automakers and research institutions typically replace actual road testing with the construction of high-precision electric vehicle powertrain test benches or chassis dynamometer systems. This allows for functional testing and performance evaluation of electric vehicle powertrain systems, control strategies, and energy management within the laboratory. This approach not only ensures the safety of researchers but also significantly reduces the development costs and shortens the development cycle of electric vehicles, while simultaneously improving the accuracy and repeatability of test data.

[0003] Due to cost and structural limitations, the inertia of the chassis dynamometer rollers is much smaller than the actual driving inertia of the electric vehicle under test. Therefore, electrical inertia compensation technology is needed to simulate the vehicle's dynamics on real roads. While the applicant's earlier application, CN105569930 A, achieved good results on a wind turbine test bench with relatively stable mechanical transmission (mostly direct drive or gearbox connection), unlike the wind turbine test bench, the electric vehicle test bench suffers from unavoidable high-frequency noise caused by inverter PWM switching, tire-roller contact slippage, and mechanical resonance. If the acceleration acquisition method of this patent is directly applied to an electric vehicle test bench under high-inertia simulation conditions, its pure differential circuit will drastically amplify the aforementioned high-frequency noise, leading to severe high-frequency fluctuations in the compensation torque. This significantly reduces the accuracy of the speed simulation and distorts the experimental data, failing to meet the requirements for high-precision dynamic performance testing of electric vehicles. Summary of the Invention

[0004] The purpose of this invention is to address the problem that existing inertia compensation strategies amplify inherent high-frequency noise in the system under high-multiple inertia simulation conditions, resulting in insufficient speed simulation accuracy. This invention proposes a mechanical inertia compensation method for electric vehicle test benches for high-precision dynamic simulation.

[0005] The technical solution to achieve the objective of this invention is as follows: On the one hand, a method for compensating the mechanical inertia of an electric vehicle test bench for high-precision dynamic simulation is provided, the method comprising the following steps:

[0006] Step 1: Determine the actual electric vehicle model that the chassis dynamometer needs to simulate and calculate its equivalent moment of inertia. Moment of inertia of the dynamometer drum itself And calculate the inertia compensation coefficient based on the relationship between the two. At the same time, determine the system sampling period. ;

[0007] Step 2: In the inertia compensator, a second-order extended state observer (ESO) is used instead of the traditional differential element using the finite difference method. This second-order ESO expands the total system disturbance, including acceleration dynamics and unmodeled dynamics, into new state variables for unified observation, thereby obtaining a high-precision acceleration estimate for inertia compensation. ;

[0008] Step 3: Substitute the inertia compensator from Step 2 into the closed-loop control loop of the electric vehicle power test bench, establish a discrete domain model of the closed-loop system containing the inertia compensator, and construct the characteristic equations of the closed-loop system. , where the coefficient , , , Bandwidth including ESO Filter coefficients Inertia compensation coefficient and system sampling period ;

[0009] Step 4: For the characteristic equation, apply the necessary and sufficient conditions of the Jury criterion to calculate the specific filter coefficients. Inertia compensation coefficient and system sampling period Below, ensure the bandwidth of the system closed loop is stable. The theoretical boundary range;

[0010] Step 5, in bandwidth Within the theoretical boundary range, the Sparrow Search Algorithm (SSA) with an adaptive boundary warning mechanism is used to evaluate the parameters. Optimization is performed; specifically, this includes using the time-weighted absolute error integral criterion and a noise penalty term. Using the objective function, a dynamic anti-predation mechanism based on the operating state is established using a reconnaissance and early warning system: a dynamic safety threshold is constructed based on the real-time rotation speed and acceleration of the experimental platform. The safety threshold is reduced under low-speed or stable operating conditions to allow for extreme optimization, and the safety threshold is increased under high-speed or rapid acceleration and deceleration operating conditions to reserve a stability margin, thereby achieving the best balance between robustness and tracking accuracy under all operating conditions.

[0011] Furthermore, the inertia compensation coefficient mentioned in step 1 The specific calculation formula is as follows:

[0012] .

[0013] Furthermore, the discrete-domain state update equation for the second-order ESO in step 2 is as follows:

[0014]

[0015] In the formula, This is the speed observation error, used to correct the state estimate at the next moment. for The actual rotational speed of the experimental platform rollers, collected by the time sensor; This is an estimated value for the rotational speed. This is an estimate of the total system disturbance, including acceleration dynamics; the observer gain is configured as follows: ; , They are respectively time, The estimated rotational speed output by the time-of-flight observer. , They are respectively time, The estimated total system disturbance output by the time-of-flight observer; The system sampling period is To control the gain, for The electromagnetic torque output by the motor at all times. The bandwidth of a second-order ESO.

[0016] Furthermore, the acceleration estimate in step 2 Depend on The result is obtained after processing with a first-order digital filter.

[0017] Furthermore, in step 3, the coefficient , , , They are represented as follows:

[0018] .

[0019] Furthermore, the necessary and sufficient condition for the Jury criterion in step 4 is that the following four inequality constraints must be satisfied simultaneously:

[0020] .

[0021] Furthermore, in step 4, the bandwidth is calculated. The theoretical boundary range is specifically: in The system of inequalities that provide sufficient and necessary conditions on a two-dimensional coordinate plane has a system of overlapping solution sets, which constitute the theoretical boundary region.

[0022] Furthermore, in step 5, the dynamic security threshold... The functional relationship between the experimental platform and its real-time operating status is as follows:

[0023]

[0024] In the formula, To ensure a minimum stability margin, the basic safety threshold, and These represent the maximum rotational speed and maximum acceleration designed for the experimental platform. and These respectively characterize the speed-related risk factors and the acceleration-related risk factors; , These represent the real-time rotational speed and acceleration of the experimental platform, respectively.

[0025] Furthermore, in step 5, during the iterative process of the sparrow search algorithm, the scout / early warning agent calculates in real time the Euclidean distance from its own position to the Jury stable boundary obtained in step 4. And compared with the dynamic security threshold calculated at the current moment. Compare, and once a decision is made This triggers anti-predation behavior, forcing individuals to abandon their current position and move closer to the global optimal position of the population, so as to achieve an adaptive balance between system stability margin and dynamic tracking accuracy across all operating conditions.

[0026] On the other hand, a mechanical inertia compensation system for an electric vehicle test bench for high-precision dynamic simulation is provided, the system comprising:

[0027] The first module is used to: determine the actual electric vehicle model that the chassis dynamometer needs to simulate, and calculate its equivalent moment of inertia. Moment of inertia of the dynamometer drum itself And calculate the inertia compensation coefficient based on the relationship between the two. At the same time, determine the system sampling period. ;

[0028] The second module is used to replace the traditional differential element using the finite difference method in the inertia compensator with a second-order extended state observer (ESO). This second-order ESO expands the total system disturbance, including acceleration dynamics and unmodeled dynamics, into new state variables for unified observation, obtaining a high-precision acceleration estimate for inertia compensation. ;

[0029] The third module is used to: substitute the inertia compensator into the closed-loop control loop of the electric vehicle power test bench, establish a discrete domain model of the closed-loop system containing the inertia compensator, and construct the characteristic equations of the closed-loop system. , where the coefficient , , , Bandwidth including ESO Filter coefficients Inertia compensation coefficient and system sampling period ;

[0030] The fourth module is used to: apply the necessary and sufficient conditions of the Jury criterion to the characteristic equation, and calculate the specific filter coefficients. Inertia compensation coefficient and system sampling period Below, ensure the bandwidth of the system closed loop is stable. The theoretical boundary range;

[0031] The fifth module is used to implement: in bandwidth Within the theoretical boundary range, a sparrow search algorithm with an adaptive boundary warning mechanism for operating conditions is used to analyze the parameters. Optimization is performed; specifically, this includes using the time-weighted absolute error integral criterion and a noise penalty term. Using the objective function, a dynamic anti-predation mechanism based on the operating state is established using a reconnaissance and early warning system: a dynamic safety threshold is constructed based on the real-time rotation speed and acceleration of the experimental platform. The safety threshold is reduced under low-speed or stable operating conditions to allow for extreme optimization, and the safety threshold is increased under high-speed or rapid acceleration and deceleration operating conditions to reserve a stability margin, thereby achieving the best balance between robustness and tracking accuracy under all operating conditions.

[0032] Compared with the prior art, the significant advantages of this invention are:

[0033] (1) The present invention uses ESO to achieve high-precision disturbance suppression and acceleration extraction. The observer uses the total disturbance, including acceleration and unmodeled dynamics, as a state variable for observation. While accurately extracting the compensation signal, it effectively isolates the high-frequency noise generated by the inverter PWM switch and mechanical contact, and significantly reduces the speed tracking deviation.

[0034] (2) This invention uses a closed-loop discrete model and the Jury criterion to quantitatively calculate the parameter stability boundary. This provides a basis for subsequent control parameter tuning and ensures that the parameter design is always within the theoretical stability domain under high-inertia simulation conditions.

[0035] (3) The present invention adopts the Sparrow Search Algorithm (SSA) with an adaptive boundary warning mechanism. This method breaks through the limitation of the single objective function of traditional optimization algorithms, and constructs a dynamic safety threshold based on real-time rotational speed and acceleration, so that the system can maintain the best balance between dynamic response speed and anti-disturbance capability in the entire operating range.

[0036] (4) This invention achieves the simulation of electric vehicles with large inertia while ensuring the accuracy of dynamic simulation, thereby reducing the hardware investment and R&D costs of electric vehicle power system test platform.

[0037] The present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0038] Figure 1 This is a flowchart illustrating the overall implementation of a mechanical inertia compensation method for an electric vehicle test bench for high-precision dynamic simulation in one embodiment.

[0039] Figure 2 This is a control block diagram of an electric vehicle dynamic simulation experimental platform system with an inertia compensator in one embodiment.

[0040] Figure 3 This is a control block diagram of a discrete model of an electric vehicle power closed-loop system with an inertia compensator constructed in one embodiment.

[0041] Figure 4 In one embodiment, during the system sampling period =0.001s, filter coefficient Below, the ESO bandwidth is calculated based on the Jury criterion. With inertia compensation factor The theoretical stability boundary.

[0042] Figure 5 This is a flowchart of parameter optimization for a sparrow search algorithm based on a working condition adaptive boundary early warning mechanism in one embodiment.

[0043] Figure 6 This is a comparison of the speed tracking performance of the method of the present invention and the traditional differential method on an electric vehicle test bench under the condition that the optimal control parameters are determined by the parameter optimization method in step 5 in one embodiment. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0045] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0046] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0047] In one embodiment, combined Figure 1 This paper provides a method for compensating the mechanical inertia of an electric vehicle test bench for high-precision dynamic simulation. The method includes the following steps:

[0048] Step 1: Determine the actual electric vehicle model that the chassis dynamometer needs to simulate and calculate its equivalent moment of inertia. Moment of inertia of the dynamometer drum itself And calculate the inertia compensation coefficient based on the relationship between the two. At the same time, determine the system sampling period. ;

[0049] Step 2: In the inertia compensator, a second-order extended state observer (ESO) is used to replace the traditional differential element using the finite difference method. This second-order ESO expands the total system disturbance, including acceleration dynamics and unmodeled dynamics, into new state variables for unified observation, obtaining high-precision acceleration estimates for inertia compensation. ;

[0050] Step 3: Substitute the inertia compensator from Step 2 into the closed-loop control loop of the electric vehicle power test bench, establish a discrete domain model of the closed-loop system containing the inertia compensator, and construct the characteristic equations of the closed-loop system. , where the coefficient , , , Bandwidth including ESO Filter coefficients Inertia compensation coefficient and system sampling period ;

[0051] Step 4: For the characteristic equation, apply the necessary and sufficient conditions of the Jury criterion to calculate the specific filter coefficients. Inertia compensation coefficient and system sampling period Below, ensure the bandwidth of the system closed loop is stable. The theoretical boundary range;

[0052] Step 5, in bandwidth Within the theoretical boundary range, the Sparrow Search Algorithm (SSA) with an adaptive boundary warning mechanism is used to analyze the parameters. Optimization is performed, specifically including: using the time-weighted absolute error integral (ITAE) criterion and a noise penalty term. Using the objective function, a dynamic anti-predation mechanism based on the operating state is established using a reconnaissance and early warning system: a dynamic safety threshold is constructed based on the real-time rotation speed and acceleration of the experimental platform. The safety threshold is reduced under low-speed or stable operating conditions to allow for extreme optimization, and the safety threshold is increased under high-speed or rapid acceleration and deceleration operating conditions to reserve a stability margin, thereby achieving the best balance between robustness and tracking accuracy under all operating conditions.

[0053] Furthermore, in one embodiment, step 1 aims to clarify the physical characteristics of the controlled object of the chassis dynamometer system and the target operating parameters to be simulated, providing a numerical basis for the subsequent controller design. The specific process of step 1 includes:

[0054] Step 1-1: Calculate the equivalent moment of inertia of the actual electric vehicle model on the chassis dynamometer rollers. Moment of inertia of the chassis dynamometer roller Both can be determined by the following formula:

[0055]

[0056] in, For the quality of electric vehicles, For the mass of the chassis dynamometer roller, The radius of the dynamometer roller on the chassis.

[0057] Step 1-2, calculate the inertia compensation coefficient. It can be determined by the following formula:

[0058]

[0059] This invention is mainly aimed at Simulation of high-magnitude inertia with relatively large values. Under such conditions, due to... As the value increases, the compensation gain in the control loop will be significantly improved, causing the traditional differential element to further amplify the inherent high-frequency noise of the system.

[0060] Furthermore, in one embodiment, step 2 aims to design an accelerometer with strong resistance to high-frequency noise, replacing the pure differential element in traditional inertia compensation strategies that is extremely sensitive to noise. Step 2 specifically includes:

[0061] Step 2-1: Establish a second-order ESO continuous domain model:

[0062] First, a second-order extended state-space model is established for the dynamic system of the roller on the electric vehicle test bench. The system state variables are defined as follows: Let... Drum speed ,make The total system disturbance is defined as the set of acceleration dynamics required for vehicle inertia simulation, as well as dynamics not modeled on the test bench (such as inverter PWM switching, tire-roller contact slippage, and high-frequency noise caused by mechanical resonance).

[0063] Based on the principle of active disturbance rejection control, the discrete-time second-order ESO state-observation equation is constructed as follows:

[0064]

[0065] in, The actual rotational speed measured by the sensor; To control the drum speed The estimated value; To the total disturbance of the system The estimated value; This is the observation error; This is the observer gain.

[0066] To reduce the number of parameters to be tuned, the bandwidth method is used to configure the observer poles. The observer characteristic equation is configured as follows: The parameter relationships in the continuous domain can be obtained as follows:

[0067]

[0068] Step 2-2, Discretization of the second-order ESO model:

[0069] To apply the second-order ESO model established in step 2-1 to the digital controller of the electric vehicle test bench, it needs to be discretized. Considering the real-time requirements and computational resource limitations of the test bench control system, this step uses the Euler forward difference method to discretize the continuous domain equations. The sampling period of the digital control system is set to... The current sampling time is The derivative terms in the continuous domain equations Approximately in difference form After simplification, the discrete-time state update equation of the second-order ESO model is as follows:

[0070]

[0071] In the formula, for The actual rotational speed of the experimental platform rollers, collected by the time sensor; for The electromagnetic torque output by the motor at all times; for The estimated rotational speed output by the time-of-flight observer; for The estimated total system disturbance output by the time-of-flight observer. This variable contains the acceleration dynamics required for vehicle inertia simulation, as well as unmodeled dynamic disturbances; The rotational speed observation error is used to correct the state estimate at the next moment; , They are respectively time, The estimated total system disturbance output by the time-of-flight observer; The system sampling period is To control the gain (in order to extract the acceleration signal for compensation and to avoid introducing coupling interference from the control quantity, it is not currently introduced in ESO), feedforward, i.e. ), for The electromagnetic torque output by the motor at all times. The bandwidth of a second-order ESO.

[0072] Furthermore, in one embodiment, step 3 aims to incorporate the inertia compensator into the closed-loop control system of the electric vehicle power test bench, and, in conjunction with the characteristics of the filter and the controlled object, construct a complete closed-loop mathematical model to provide a theoretical basis for subsequent stability analysis. The specific implementation process is as follows:

[0073] Step 3-1: Derive the discrete-domain transfer function of the second-order ESO. :

[0074] Based on the discrete state update equation established in step 2-2, the error feedback mechanism in the observer is analyzed. Transformation analysis. Assume the input is the drum speed. The output is the total disturbance estimate. Without considering control quantities In the case of feedforward influence, the transmission relationship between the two is derived from the state equation:

[0075]

[0076] Eliminating intermediate variables by combining the above system of equations The discrete-domain transfer function of the second-order ESO from the speed input to the disturbance estimation output is obtained by rearranging. :

[0077]

[0078] Step 3-2, construct the characteristic equations of the closed-loop discrete system of the electric vehicle test bench:

[0079] The discrete-domain transfer function of the discretized second-order ESO described above Substitute this into the inertia compensation loop of the electric vehicle power test bench. Derive the open-loop transfer function of the closed-loop discrete system of the electric vehicle test bench. :

[0080] Finally, according to discrete control theory, the stability of a closed-loop system depends on its characteristic equation. .Will Substitute and remove the denominator to simplify, and then... Arrange the powers in descending order, and sort them to obtain the information about The standard cubic characteristic equation yields:

[0081]

[0082] Among them, the coefficients of each order , , , Bandwidth including ESO Filter coefficients Inertia compensation coefficient and system sampling period The specific expression is as follows:

[0083] .

[0084] Furthermore, in one embodiment, step 4 aims to use discrete system stability criteria to quantitatively analyze the third-order closed-loop model established in step 3, thereby determining the safe parameter range that can guarantee the stable operation of the experimental system. The specific implementation process is as follows:

[0085] Step 4-1, construct the Jury stability criterion for the third-order system:

[0086] According to discrete control theory, for a third-order system, the necessary and sufficient condition for the system's closed-loop stability is that the following four inequality constraints must be satisfied simultaneously:

[0087]

[0088] Step 4-2, calculate the parameter stability boundary:

[0089] The coefficient expression known in step 3-2 , , , Substitute these into the four inequalities above to solve:

[0090] Condition ①:

[0091] when , , When, condition ① always holds true.

[0092] Condition 2:

[0093] when At that time, condition ② is met.

[0094] Condition ③:

[0095] when At that time, condition ③ is met.

[0096] Condition 4:

[0097] when At that time, condition ④ is met.

[0098] exist Solving this system of inequalities in a two-dimensional coordinate plane, the region of coincident solution sets represents the theoretical stability boundary. This stability boundary is only valid when the chosen parameter combination... Only when the system falls within this enclosed area can it be theoretically guaranteed that the small inertia chassis dynamometer is stable when simulating the dynamics of a large inertia electric vehicle.

[0099] Furthermore, in one embodiment, step 5 aims to resolve the multi-objective constraint conflict between the high dynamic tracking accuracy requirement under low-speed conditions and the strong system robustness requirement under high-speed conditions in high-inertia simulation. By introducing the population co-evolution mechanism of the sparrow search algorithm and combining it with real-time operating status feedback from the experimental platform, an adaptive global optimization strategy is constructed. The specific implementation process is as follows:

[0100] Step 5-1: Construct a composite objective function that balances steady-state tracking accuracy and high-frequency noise suppression.

[0101] To minimize the steady-state tracking error and find the optimal control parameters to suppress the impact of high-frequency noise on the system, this step employs the time-weighted absolute error integral (ITAE) criterion and a noise penalty term. Constructing the fitness function :

[0102]

[0103] The first item utilizes time weighting. The steady-state residual error is heavily penalized, forcing the algorithm to find the parameter with the fastest convergence speed; the second term This is a high-frequency noise penalty term. These are the weighting coefficients. Fitness function. The smaller the value, the higher the tracking accuracy and the smoother the waveform.

[0104] Step 5-2, Population Initialization and Role Assignment:

[0105] The theoretical stability boundary determined in step 4 ( Figure 5 Within the blue area, this step uses chaotic mapping to randomly generate... The initial position of the sparrow Based on fitness values, the top 20% of individuals are selected as discoverers, the remaining 80% are selected as joiners, and 10% to 20% of individuals are randomly selected from the population to serve as scouts and early warning agents.

[0106] Step 5-3: Perform position update based on theoretical stability boundary:

[0107] This step, based on the population cooperation mechanism of the sparrow search algorithm, implements differentiated position update strategies for discoverers and joiners to achieve synergy between global exploration and local development. First, the discoverer, as the core of the population, is responsible for conducting a global search within the theoretical stability boundary, and its position update follows the mathematical model below:

[0108]

[0109] in, This represents the current iteration number. The maximum number of iterations, The j-th dimension parameter representing the i-th individual; It is a random number. and These are the warning value and the safety threshold, respectively. The numbers are random numbers that follow a normal distribution. Their position update formula includes a long-distance jump factor, enabling them to quickly traverse high-fitness regions and effectively avoid local optima.

[0110] The joiner then performs a local follow-search based on the discoverer's location, and its location update is described as follows:

[0111]

[0112] in, This is the optimal position currently occupied by the population. The worst position This is the search direction matrix. When... When, the participants perform refined optimization near the optimal position; when When a new participant joins, it performs a random traversal. A dynamic competition mechanism is also introduced: if a new participant's calculated fitness is better than the current discoverer's, it immediately replaces the discoverer, ensuring the algorithm achieves a deep approximation of the high-precision parameter region.

[0113] Step 5-4: Execute adaptive boundary warning based on operating condition awareness:

[0114] The core of this step is to introduce an adaptive safety threshold model based on operating conditions to solve the problem that fixed thresholds cannot simultaneously achieve high accuracy at low speeds and high robustness at high speeds.

[0115] Considering the increased mechanical vibration at high speeds and the enhanced nonlinear characteristics during rapid acceleration and deceleration, the system is more prone to approaching the instability boundary. Therefore, a dynamic safety threshold is defined. Regarding the real-time rotation speed of the experimental platform With acceleration Functions:

[0116]

[0117] in, To ensure a minimum stability margin, the basic safety threshold, and These represent the maximum rotational speed and maximum acceleration designed for the experimental platform. and These respectively characterize the speed-related risk factors and the acceleration-related risk factors, which together determine the stability margin requirements under different operating conditions.

[0118] During the iterative process of the sparrow search algorithm, the scout calculates in real time the Euclidean distance from its own position to the Jury stability boundary described in step 4. and the result calculated at the current time Comparison. Under low-speed or steadily changing, low-risk operating conditions, the threshold... Automatic contraction approximation This allows the population to optimize within a high-gain parameter space approaching the stability boundary to obtain the theoretically optimal dynamic tracking accuracy; however, under high-speed or rapidly changing high-risk conditions, the threshold... The automatic and significant increase allows the population to perform optimization searches within a safe parameter space far from the instability boundary, thus reserving sufficient system stability margin. Once determined... This triggers anti-predation behavior, forcing the individual to abandon its current position and move towards the global optimal position of the population, thereby achieving an adaptive balance between system stability margin and dynamic tracking accuracy across all operating conditions.

[0119] Step 5-5, output the optimal solution:

[0120] Maximum number of iterations Then, output the coordinates of the sparrow with the smallest fitness value. This is the final determined optimal control parameter.

[0121] This method fully considers the unavoidable interferences in electric vehicle test benches, such as inherent high-frequency slip noise and mechanical resonance generated by inverter PWM switching and tire-roller contact. An extended state observer is introduced into the inertia compensator to replace the traditional differential element. This observer expands the total system disturbance, including acceleration dynamics and unmodeled dynamics, into a new state variable for unified observation and compensation, solving the problem of decreased speed simulation accuracy caused by high-frequency noise amplification under high-compensation conditions. Based on this, a closed-loop discrete mathematical model of the system containing an ESO is constructed, and its characteristic equation and Jury criterion are used to determine the stability boundary of the system under high-compensation conditions. Finally, a sparrow search algorithm with an adaptive boundary warning mechanism is used to optimize the observer gain within the stability domain. This algorithm constructs a dynamic safety threshold based on the real-time speed and acceleration of the test bench, and resolves the conflict between the requirements for high accuracy at low speeds and strong robustness at high speeds by adaptively adjusting the stability margin constraints during optimization. This invention effectively solves the interference of high-frequency noise on the compensation loop, achieving high-precision and high-stability simulation of the mechanical dynamics of large-inertia electric vehicles under all operating conditions.

[0122] In one embodiment, a mechanical inertia compensation system for an electric vehicle test bench for high-precision dynamic simulation is provided, the system comprising:

[0123] The first module is used to: determine the actual electric vehicle model that the chassis dynamometer needs to simulate, and calculate its equivalent moment of inertia. Moment of inertia of the dynamometer drum itself And calculate the inertia compensation coefficient based on the relationship between the two. At the same time, determine the system sampling period. ;

[0124] The second module is used to replace the traditional differential element using the finite difference method in the inertia compensator with a second-order extended state observer (ESO). This second-order ESO expands the total system disturbance, including acceleration dynamics and unmodeled dynamics, into new state variables for unified observation, obtaining a high-precision acceleration estimate for inertia compensation. ;

[0125] The third module is used to: substitute the inertia compensator into the closed-loop control loop of the electric vehicle power test bench, establish a discrete domain model of the closed-loop system containing the inertia compensator, and construct the characteristic equations of the closed-loop system. , where the coefficient , , , Bandwidth including ESO Filter coefficients Inertia compensation coefficient and system sampling period ;

[0126] The fourth module is used to: apply the necessary and sufficient conditions of the Jury criterion to the characteristic equation, and calculate the specific filter coefficients. Inertia compensation coefficient and system sampling period Below, ensure the bandwidth of the system closed loop is stable. The theoretical boundary range;

[0127] The fifth module is used to implement: in bandwidth Within the theoretical boundary range, a sparrow search algorithm with an adaptive boundary warning mechanism for operating conditions is used to analyze the parameters. Optimization is performed; specifically, this includes using the time-weighted absolute error integral criterion and a noise penalty term. Using the objective function, a dynamic anti-predation mechanism based on the operating state is established using a reconnaissance and early warning system: a dynamic safety threshold is constructed based on the real-time rotation speed and acceleration of the experimental platform. The safety threshold is reduced under low-speed or stable operating conditions to allow for extreme optimization, and the safety threshold is increased under high-speed or rapid acceleration and deceleration operating conditions to reserve a stability margin, thereby achieving the best balance between robustness and tracking accuracy under all operating conditions.

[0128] Specific limitations regarding the mechanical inertia compensation system for electric vehicle test benches for high-precision dynamic simulation can be found in the limitations of the mechanical inertia compensation method for electric vehicle test benches for high-precision dynamic simulation described above, and will not be repeated here. Each module in the aforementioned mechanical inertia compensation system for electric vehicle test benches for high-precision dynamic simulation can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0129] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements:

[0130] Step 1: Determine the actual electric vehicle model that the chassis dynamometer needs to simulate and calculate its equivalent moment of inertia. Moment of inertia of the dynamometer drum itself And calculate the inertia compensation coefficient based on the relationship between the two. At the same time, determine the system sampling period. ;

[0131] Step 2: In the inertia compensator, a second-order extended state observer (ESO) is used to replace the traditional differential element using the finite difference method. This second-order ESO expands the total system disturbance, including acceleration dynamics and unmodeled dynamics, into new state variables for unified observation, obtaining high-precision acceleration estimates for inertia compensation. ;

[0132] Step 3: Substitute the inertia compensator from Step 2 into the closed-loop control loop of the electric vehicle power test bench, establish a discrete domain model of the closed-loop system containing the inertia compensator, and construct the characteristic equations of the closed-loop system. , where the coefficient , , , Bandwidth including ESO Filter coefficients Inertia compensation coefficient and system sampling period ;

[0133] Step 4: For the characteristic equation, apply the necessary and sufficient conditions of the Jury criterion to calculate the specific filter coefficients. Inertia compensation coefficient and system sampling period Below, ensure the bandwidth of the system closed loop is stable. The theoretical boundary range;

[0134] Step 5, in bandwidth Within the theoretical boundary range, the Sparrow Search Algorithm (SSA) with an adaptive boundary warning mechanism is used to analyze the parameters. Optimization is performed, specifically including: using the time-weighted absolute error integral (ITAE) criterion and a noise penalty term. Using the objective function, a dynamic anti-predation mechanism based on the operating state is established using a reconnaissance and early warning system: a dynamic safety threshold is constructed based on the real-time rotation speed and acceleration of the experimental platform. The safety threshold is reduced under low-speed or stable operating conditions to allow for extreme optimization, and the safety threshold is increased under high-speed or rapid acceleration and deceleration operating conditions to reserve a stability margin, thereby achieving the best balance between robustness and tracking accuracy under all operating conditions.

[0135] For specific limitations on each step, please refer to the limitations on the mechanical inertia compensation method for electric vehicle test benches for high-precision dynamic simulation mentioned above, which will not be repeated here.

[0136] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program being implemented when executed by a processor:

[0137] Step 1: Determine the actual electric vehicle model that the chassis dynamometer needs to simulate and calculate its equivalent moment of inertia. Moment of inertia of the dynamometer drum itself And calculate the inertia compensation coefficient based on the relationship between the two. At the same time, determine the system sampling period. ;

[0138] Step 2: In the inertia compensator, a second-order extended state observer (ESO) is used to replace the traditional differential element using the finite difference method. This second-order ESO expands the total system disturbance, including acceleration dynamics and unmodeled dynamics, into new state variables for unified observation, obtaining high-precision acceleration estimates for inertia compensation. ;

[0139] Step 3: Substitute the inertia compensator from Step 2 into the closed-loop control loop of the electric vehicle power test bench, establish a discrete domain model of the closed-loop system containing the inertia compensator, and construct the characteristic equations of the closed-loop system. , where the coefficient , , , Bandwidth including ESO Filter coefficients Inertia compensation coefficient and system sampling period ;

[0140] Step 4: For the characteristic equation, apply the necessary and sufficient conditions of the Jury criterion to calculate the specific filter coefficients. Inertia compensation coefficient and system sampling period Below, ensure the bandwidth of the system closed loop is stable. The theoretical boundary range;

[0141] Step 5, in bandwidth Within the theoretical boundary range, the Sparrow Search Algorithm (SSA) with an adaptive boundary warning mechanism is used to analyze the parameters. Optimization is performed, specifically including: using the time-weighted absolute error integral (ITAE) criterion and a noise penalty term. Using the objective function, a dynamic anti-predation mechanism based on the operating state is established using a reconnaissance and early warning system: a dynamic safety threshold is constructed based on the real-time rotation speed and acceleration of the experimental platform. The safety threshold is reduced under low-speed or stable operating conditions to allow for extreme optimization, and the safety threshold is increased under high-speed or rapid acceleration and deceleration operating conditions to reserve a stability margin, thereby achieving the best balance between robustness and tracking accuracy under all operating conditions.

[0142] For specific limitations on each step, please refer to the limitations on the mechanical inertia compensation method for electric vehicle test benches for high-precision dynamic simulation mentioned above, which will not be repeated here.

[0143] As a specific example, the invention will be further verified and illustrated in one embodiment.

[0144] This embodiment verifies the effectiveness of the electric vehicle test bench mechanical inertia compensation method for high-precision dynamic simulation of the present invention, based on... Figure 1 The overall implementation flowchart shown illustrates the construction of a hardware-in-the-loop (HIL) dynamic simulation experimental platform for electric vehicles. The platform's main equipment includes one test vehicle (electric vehicle), one dual-roller chassis dynamometer, one host computer responsible for offline algorithm optimization and global status monitoring, and one programmable logic controller (PLC) serving as the lower-level main controller. Its system control block diagram is shown below. Figure 2 As shown.

[0145] To suppress the inherent high-frequency noise of the electric vehicle test bench, this invention introduces a second-order extended state observer and adds it to the control block diagram of the closed-loop discrete domain model of the electric vehicle power system containing an inertia compensator, as shown in the figure. Figure 3 As shown in Table 1, the main physical platform implementation uses the MATLAB / Simulink software platform to perform theoretical stability boundary calculations for the third-order Jury criterion and optimization using the Sparrow Search Algorithm (SSA) with an adaptive early warning mechanism. The lower-level PLC uses the TwinCAT3 real-time kernel to execute the second-order ESO discretized control program and inertia compensation algorithm, and sends torque commands to the dynamometer's converter in real-time via the high-speed EtherCAT bus. The main physical parameters of the electric vehicle dynamic simulation experimental platform are shown in Table 1.

[0146] Table 1 Main physical parameters of the electric vehicle dynamic simulation test platform

[0147]

[0148] Based on the system physical parameters shown in Table 1, experimental test conditions for high-magnification inertia simulation were further constructed: the inertia compensation coefficient was set. The system's initial rotational speed was set to 49 (the inertia of the vehicle under test is 50 times that of the roller) to simulate the dynamics of a high-inertia electric vehicle. The initial rotational speed was stabilized at 1200 rpm. A speed step command with a target value of 3000 rpm is applied at all times.

[0149] Under this condition, the theoretical stability boundary range is first calculated based on the Jury criterion, such as... Figure 4 As shown (the blue area represents the theoretically stable region of the system). Based on this, in order to find the optimal solution that balances accuracy and robustness within the aforementioned stability region, this embodiment performs the following... Figure 5 The diagram illustrates the parameter optimization process of the sparrow search algorithm based on an adaptive boundary warning mechanism. This process introduces a dynamic safety threshold that varies with the real-time rotational speed and acceleration of the experimental setup to adaptively constrain and iterate the population, thereby determining the optimal control parameters for the ESO. And will adopt an improved method with optimal control parameters and filter parameters. The traditional differential method was compared and controlled on a physical test bench, and the final speed tracking results were as follows: Figure 6 As shown in the figure, the experimental comparison results show that when a speed step command is issued, due to the characteristics of the traditional differential element, the speed trajectory using the traditional differential method (blue curve) contains a large amount of high-frequency noise, which seriously affects the control quality. In contrast, the speed trajectory of the method of this invention (orange curve) is smooth and stable. While achieving fast and steady-state error-free tracking of speed step commands, it eliminates high-frequency noise interference in the waveform, significantly reducing the steady-state speed standard deviation by 60.95%, verifying that the method of this invention has excellent noise suppression capability and steady-state accuracy while ensuring high dynamic response speed.

[0150] In summary, the proposed method for mechanical inertia compensation on an electric vehicle test bench for high-precision dynamic simulation firstly utilizes an extended state observer to expand the total system disturbance, including acceleration dynamics and unmodeled dynamics, into new state variables for unified observation and compensation, thus solving the problem of amplified high-frequency noise on the test bench leading to decreased speed simulation accuracy under high-inertia simulation. Secondly, this invention combines the discrete-domain Jury stability criterion to quantify the system stability boundary under high-inertia simulation. Finally, within this stability domain, a sparrow search algorithm with an adaptive boundary warning mechanism is introduced. By dynamically adjusting the safety threshold during the optimization process, the contradiction between low-speed high precision and high-speed strong robustness is resolved, ultimately achieving high-precision simulation of the mechanical dynamics of large-inertia vehicles under all operating conditions.

[0151] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention without departing from its spirit and scope should be included within the protection scope of the present invention.

Claims

1. A method for compensating the mechanical inertia of an electric vehicle test bench for high-precision dynamic simulation, characterized in that, The method includes the following steps: Step 1: Determine the actual electric vehicle model that the chassis dynamometer needs to simulate and calculate its equivalent moment of inertia. Moment of inertia of the dynamometer drum itself And calculate the inertia compensation coefficient based on the relationship between the two. At the same time, determine the system sampling period. ; Step 2: In the inertia compensator, a second-order extended state observer (ESO) is used to replace the traditional differential element using the finite difference method. This second-order ESO expands the total system disturbance, including acceleration dynamics and unmodeled dynamics, into new state variables for unified observation, obtaining high-precision acceleration estimates for inertia compensation. ; Step 3: Substitute the inertia compensator from Step 2 into the closed-loop control loop of the electric vehicle power test bench, establish a discrete domain model of the closed-loop system containing the inertia compensator, and construct the characteristic equations of the closed-loop system. , where the coefficient , , , Bandwidth including ESO Filter coefficients Inertia compensation coefficient and system sampling period ; Step 4: For the characteristic equation, apply the necessary and sufficient conditions of the Jury criterion to calculate the specific filter coefficients. Inertia compensation coefficient and system sampling period Below, ensure the bandwidth of the system closed loop is stable. The theoretical boundary range; Step 5, in bandwidth Within the theoretical boundary range, a sparrow search algorithm with an adaptive boundary warning mechanism for operating conditions is used to analyze the parameters. Optimization is performed; specifically, this includes using the time-weighted absolute error integral criterion and a noise penalty term. As the objective function, a dynamic anti-predation mechanism based on the operating state is established using a reconnaissance and early warning system: a dynamic safety threshold is constructed based on the real-time rotation speed and acceleration of the test bench. The safety threshold is reduced under low-speed or stable conditions to allow for extreme optimization, and the safety threshold is increased under high-speed or rapid acceleration and deceleration conditions to reserve a stability margin, thereby achieving the best balance between robustness and tracking accuracy under all operating conditions. The inertia compensation coefficient mentioned in step 1 The specific calculation formula is as follows: ； In step 3, the coefficient , , , They are represented as follows: ； Observer gain configured as .

2. The method for compensating the mechanical inertia of an electric vehicle test bench for high-precision dynamic simulation according to claim 1, characterized in that, The discrete-domain state update equation for the second-order ESO in step 2 is as follows: ； In the formula, This is the speed observation error, used to correct the state estimate at the next moment. for The actual rotational speed of the experimental platform rollers, collected by the time sensor; This is an estimated value for the rotational speed. This is an estimate of the total system disturbance, including acceleration dynamics; the observer gain is configured as follows: ; , They are respectively time, The estimated rotational speed output by the time-of-flight observer. , They are respectively time, The estimated total system disturbance output by the time-of-flight observer; The system sampling period is To control the gain, for The electromagnetic torque output by the motor at all times. The bandwidth of a second-order ESO.

3. The method for compensating the mechanical inertia of an electric vehicle test bench for high-precision dynamic simulation according to claim 2, characterized in that, Acceleration estimation value in step 2 Depend on The result is obtained after processing with a first-order digital filter.

4. The method for compensating the mechanical inertia of an electric vehicle test bench for high-precision dynamic simulation according to claim 1, characterized in that, In step 4, the necessary and sufficient condition for Jury's criterion is that the following four inequality constraints must be satisfied simultaneously: 。 5. The method for compensating the mechanical inertia of an electric vehicle test bench for high-precision dynamic simulation according to claim 4, characterized in that, Calculate bandwidth in step 4 The theoretical boundary range is specifically: in The system of inequalities that provide sufficient and necessary conditions on a two-dimensional coordinate plane has a system of overlapping solution sets, which constitute the theoretical boundary region.

6. The method for compensating the mechanical inertia of an electric vehicle test bench for high-precision dynamic simulation according to claim 1, characterized in that, In step 5, the dynamic security threshold The functional relationship between the experimental platform and its real-time operating status is as follows: ； In the formula, To ensure a minimum stability margin, the basic safety threshold, and These represent the maximum rotational speed and maximum acceleration designed for the experimental platform. and These respectively characterize the speed-related risk factors and the acceleration-related risk factors; , These represent the real-time rotational speed and acceleration of the experimental platform, respectively.

7. The method for compensating the mechanical inertia of an electric vehicle test bench for high-precision dynamic simulation according to claim 1, characterized in that, In step 5, during the iterative process of the sparrow search algorithm, the scout / early warning agent calculates in real time the Euclidean distance from its own position to the Jury stable boundary obtained in step 4. And compared with the dynamic security threshold calculated at the current moment. Compare, and once a decision is made This triggers anti-predation behavior, forcing individuals to abandon their current position and move closer to the global optimal position of the population, so as to achieve an adaptive balance between system stability margin and dynamic tracking accuracy across all operating conditions.

8. A mechanical inertia compensation system for an electric vehicle test bench for high-precision dynamic simulation based on the method of any one of claims 1 to 7, characterized in that, The system includes: The first module is used to: determine the actual electric vehicle model that the chassis dynamometer needs to simulate, and calculate its equivalent moment of inertia. Moment of inertia of the dynamometer drum itself And calculate the inertia compensation coefficient based on the relationship between the two. At the same time, determine the system sampling period. ; The second module is used to replace the traditional differential element using the finite difference method in the inertia compensator with a second-order extended state observer (ESO). This second-order ESO expands the total system disturbance, including acceleration dynamics and unmodeled dynamics, into new state variables for unified observation, obtaining a high-precision acceleration estimate for inertia compensation. ; The third module is used to: substitute the inertia compensator into the closed-loop control loop of the electric vehicle power test bench, establish a discrete domain model of the closed-loop system containing the inertia compensator, and construct the characteristic equations of the closed-loop system. , where the coefficient , , , Bandwidth including ESO Filter coefficients Inertia compensation coefficient and system sampling period ; The fourth module is used to: apply the necessary and sufficient conditions of the Jury criterion to the characteristic equation, and calculate the specific filter coefficients. Inertia compensation coefficient and system sampling period Below, ensure the bandwidth of the system closed loop is stable. The theoretical boundary range; The fifth module is used to implement: in bandwidth Within the theoretical boundary range, a sparrow search algorithm with an adaptive boundary warning mechanism for operating conditions is used to analyze the parameters. Optimization is performed; specifically, this includes using the time-weighted absolute error integral criterion and a noise penalty term. Using the objective function, a dynamic anti-predation mechanism based on the operating state is established using a reconnaissance and early warning system: a dynamic safety threshold is constructed based on the real-time rotation speed and acceleration of the experimental platform. The safety threshold is reduced under low-speed or stable operating conditions to allow for extreme optimization, and the safety threshold is increased under high-speed or rapid acceleration and deceleration operating conditions to reserve a stability margin, thereby achieving the best balance between robustness and tracking accuracy under all operating conditions.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 7.

Images