A control method and system for a magnetorheological damper of a vehicle suspension system

Abstract

The present invention provides a control method and system for a magnetorheological damper for a vehicle suspension system, which belongs to the field of vehicle suspension control technology, including obtaining vehicle state data to generate a vehicle state stability coefficient; obtaining road condition data on the road surface on which the vehicle is traveling to generate a road condition coefficient; generating a basic damping force of the magnetorheological damper based on the vehicle state stability coefficient and the road condition coefficient; obtaining the vehicle's historical vibration peak value and the user preference index to generate an optimized damping force of the magnetorheological damper; obtaining the real-time temperature of the magnetorheological fluid and the damper piston speed to generate a final output current value, thereby adjusting the damping force of the magnetorheological damper of the vehicle suspension system; the present invention can achieve precise control of the damping force, improve the control accuracy of the magnetorheological damper of the vehicle suspension system under complex working conditions, and optimize the vehicle's driving comfort and handling stability.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle suspension control, and in particular to a control method and system for a magnetorheological damper used in a vehicle suspension system. Background Art

[0002] As a core intelligent component of a vehicle's suspension system, magnetorheological dampers (MRDs) achieve millisecond-level dynamic damping response by adjusting the magnetic field intensity to alter the rheological properties of the MR fluid. Their performance directly impacts vehicle ride comfort and handling stability, with comfort primarily reflected in vibration reduction, while handling stability is reflected in indicators such as anti-roll capability.

[0003] The current mainstream control methods mainly rely on threshold control or fixed rule bases of vehicle state parameters, which include basic data such as lateral acceleration and roll angle rate.

[0004] However, existing technologies have significant technical limitations: First, in terms of data collection, existing solutions fail to fully integrate multi-dimensional road condition information such as road surface roughness, tire slip, road moisture content, and texture characteristic index. This results in a significant decrease in damping force control accuracy under complex operating conditions such as slippery or highly rough roads. Second, in terms of control strategy, there is a lack of a learning mechanism for user personalized preferences. In particular, it is unable to effectively record and analyze long-term data formed by users' manual adjustment habits, resulting in a systematic deviation between the output current and actual demand. These technical shortcomings make it difficult for traditional methods to achieve the optimal balance between vehicle dynamic stability and driving comfort, especially when responding to sudden road changes or meeting the needs of different driving styles. In addition, existing technologies lack the ability to dynamically compensate for real-time operating parameters such as magnetorheological fluid temperature changes and piston speed, further limiting the accuracy of damping force control. Summary of the Invention

[0005] The object of the present invention is to provide a control method and system for a magnetorheological damper for a vehicle suspension system, so as to solve at least one of the above-mentioned technical problems existing in the prior art.

[0006] To solve the above technical problems, the present invention provides a method for controlling a magnetorheological damper for a vehicle suspension system, which specifically includes the following steps:

[0007] Acquiring vehicle state data and generating a vehicle state stability coefficient; wherein the vehicle state data includes vehicle lateral acceleration, pitch angular rate, and roll angular rate;

[0008] Obtaining road condition data of the road on which the vehicle is traveling and generating a road condition coefficient; wherein the road condition data includes road surface roughness, tire slip rate, road surface moisture content, and texture characteristic index; the texture characteristic index refers to the roughness of the road surface texture;

[0009] Based on the fuzzy rule base, the basic damping force of the magnetorheological damper is generated according to the vehicle state stability coefficient and the road condition coefficient;

[0010] Based on the LSTM optimization model, the optimized damping force of the magnetorheological damper is generated according to the vehicle's historical vibration peak value and user preference index;

[0011] Obtain the real-time temperature of the magnetorheological fluid and the speed of the damper piston, establish a dynamic compensation model, and generate the final output current value;

[0012] The damping force of the magnetorheological damper of the vehicle suspension system is adjusted according to the final output current value.

[0013] On the basis of the above technical solutions, the present invention also provides the following optional technical solutions:

[0014] Further technical solution: The vehicle state stability coefficient is generated in the following manner:

[0015] By formula:

[0016] ;

[0017] Generate vehicle state stability coefficient K stab ;

[0018] In the formula, a y It represents the lateral acceleration of the vehicle, W pitch It represents the pitch angular rate, W roll It represents the roll angle rate, α, β, and γ are all weight coefficients, and α+β+γ=1.

[0019] Further technical solution: The road condition coefficient is generated in the following manner:

[0020] By formula:

[0021] ;

[0022] Generate road condition coefficient K road ;

[0023] In the formula, R rough It represents the road roughness, s represents the tire slip rate, P represents the road moisture content, M tex It represents the texture feature index, k1, k2, k3, k4 are all weight coefficients, and k1+k2+k3+k4=1.

[0024] Further technical solution: The method for generating the basic damping force of the magnetorheological damper specifically includes the following steps:

[0025] Based on the fuzzy rule base, the membership degree of the fuzzy rule set is generated according to the vehicle state stability coefficient and the road condition coefficient;

[0026] Generate rule activation strength based on the membership degree of the fuzzy rule set;

[0027] The basic damping force of the magnetorheological damper is generated according to the regular activation intensity.

[0028] Further technical solution: The membership degree of the fuzzy rule set is generated in the following manner:

[0029] ;

[0030] Generate membership degree µ of fuzzy rule set A (x);

[0031] In the formula, x represents the input value, c represents the center value of the fuzzy set, and ω represents the set width; the input value x includes the vehicle state stability coefficient and the road condition coefficient.

[0032] Further technical solution: The method for generating the rule activation strength is specifically as follows:

[0033] By formula:

[0034] ;

[0035] Generate rule activation strength g i ;

[0036] In the formula, µ Kstab Represents the vehicle state stability coefficient K stab Membership degree at the current rule level, µ Kroad Indicates the road condition coefficient K road The degree of membership in the current rule level.

[0037] Further technical solution: The basic damping force of the magnetorheological damper is generated in the following manner:

[0038] By formula:

[0039] ;

[0040] Generate the basic damping force F of the magnetorheological damper base ;

[0041] In the formula, g i It represents the rule activation strength, F i It represents the preset damping force corresponding to the i-th rule, and n represents the total number of activated rules.

[0042] Further technical solution: The method for generating the optimized damping force of the magnetorheological damper specifically includes:

[0043] Obtain vehicle historical vibration peak values ​​and user preference index;

[0044] An LSTM optimization model was established, and the vehicle's historical vibration peak value and user preference index were substituted into the LSTM model to generate the optimized damping force of the magnetorheological damper.

[0045] The LSTM model is specifically:

[0046] ;

[0047] In the formula, LSTM stands for long short-term memory network, A peak It represents the historical vibration peak value of the vehicle, U pref It represents the user preference index, F base It represents the basic damping force of the magnetorheological damper;

[0048] The user preference index U pref The specific way to obtain it is:

[0049] By formula:

[0050] ;

[0051] In the formula, U adjust It represents the latest manual adjustment value of the user, U pref （t） It represents the user preference index, U pref （t-1） It represents the user preference index of the last update, and t represents the update number.

[0052] Further technical solution: The final output current value is generated in the following manner:

[0053] Based on the damping force conversion formula, the optimized current value is generated according to the optimized damping force of the magnetorheological damper;

[0054] Obtain the real-time temperature of the magnetorheological fluid and the speed of the damper piston, establish a dynamic compensation model, and generate the final output current value;

[0055] The expression of the dynamic compensation model is specifically:

[0056] ;

[0057] In the expression, I final Indicates the final output current value, I optimal Indicates the optimized current value, K TIt represents the temperature compensation coefficient of magnetorheological fluid, T j It represents the real-time temperature of the magnetorheological fluid, T ref It represents the calibration temperature of magnetorheological fluid, K V It represents the speed compensation coefficient, v piston It represents the damper piston speed.

[0058] A control system for a magnetorheological damper of a vehicle suspension system, specifically comprising:

[0059] A vehicle state analysis unit is used to obtain vehicle state data and generate a vehicle state stability coefficient; wherein the vehicle state data includes the vehicle's lateral acceleration, pitch angular rate, and roll angular rate; the vehicle's lateral acceleration, pitch angular rate, and roll angular rate are all dimensionless data;

[0060] A road condition analysis unit is used to obtain road condition data of the road on which the vehicle is traveling and generate a road condition coefficient; the road condition data includes road surface roughness, tire slip rate, road surface moisture content, and texture characteristic index; the texture characteristic index refers to the roughness of the road surface texture; the road surface roughness, tire slip rate, road surface moisture content, and texture characteristic index are all dimensionless data;

[0061] A basic damping force analysis unit is used to generate the basic damping force of the magnetorheological damper based on the fuzzy rule base, the vehicle state stability coefficient and the road condition coefficient;

[0062] A damping force optimization unit, which generates the optimized damping force of the magnetorheological damper based on the LSTM optimization model, the vehicle's historical vibration peak value, and the user preference index;

[0063] A final output current value generating unit is used to obtain the real-time temperature of the magnetorheological fluid and the speed of the damper piston to generate a final output current value;

[0064] The control unit is used to adjust the damping force of the magnetorheological damper of the vehicle suspension system according to the final output current value.

[0065] By adopting the above technical solution, the present invention has the following beneficial effects:

[0066] The present invention integrates the vehicle state stability coefficient, multi-dimensional road condition parameters and user preference data, and combines it with a dynamic temperature and speed compensation mechanism to achieve precise regulation of the damping force, thereby improving the control accuracy of the magnetorheological damper of the vehicle suspension system under complex working conditions and optimizing the vehicle's driving comfort and handling stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0068] Figure 1 A flowchart of a method for controlling a magnetorheological damper of a vehicle suspension system provided by an embodiment of the present invention;

[0069] Figure 2 A flowchart of S3 provided in an embodiment of the present invention;

[0070] Figure 3 A flowchart of S4 provided in an embodiment of the present invention;

[0071] Figure 4 A flowchart of S5 provided in an embodiment of the present invention;

[0072] Figure 5 A schematic structural diagram of a control system for a magnetorheological damper of a vehicle suspension system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0073] The technical solutions of this application will be described clearly and completely below, in conjunction with the accompanying drawings. It should be understood that the described embodiments represent only a portion of the embodiments of this application, and not all of them. The components of this application, generally described and illustrated in the drawings herein, may be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of this application provided in the drawings is not intended to limit the scope of the claimed application, but rather merely represents selected embodiments of this application. All other embodiments derived by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application. It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. Furthermore, in the description of this application, the terms "first," "second," etc., are used solely to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0074] In existing technologies, magnetorheological dampers achieve dynamic damping force response by adjusting the magnetic field intensity to alter the rheological properties of the magnetorheological fluid. These control methods primarily rely on threshold control of vehicle state parameters or a fixed rule base. Existing technologies fail to fully integrate multidimensional road condition data, such as road moisture content or texture characteristic index, leading to inaccurate damping force control on slippery or highly rough surfaces. Furthermore, they lack a learning mechanism for user manual adjustment habits, resulting in output current offsets and difficulty in balancing the multi-objective optimization requirements of vehicle stability and ride comfort.

[0075] In order to solve the above problems, the present application proposes a control method for a magnetorheological damper of a vehicle suspension system.

[0076] like Figure 1 As shown in FIG, a method for controlling a magnetorheological damper for a vehicle suspension system provided in this embodiment specifically includes the following steps:

[0077] S1: Acquire vehicle state data and generate a vehicle state stability coefficient; wherein the vehicle state data includes the vehicle lateral acceleration, pitch angle rate, and roll angle rate;

[0078] S2: Obtaining road condition data of the road on which the vehicle is traveling and generating a road condition coefficient; wherein the road condition data includes road surface roughness, tire slip rate, road surface moisture content, and texture characteristic index; the texture characteristic index refers to the roughness of the road surface texture;

[0079] S3: Based on the fuzzy rule base, the basic damping force of the magnetorheological damper is generated according to the vehicle state stability coefficient and the road condition coefficient;

[0080] S4: Generates the optimized damping force of the magnetorheological damper based on the vehicle's historical vibration peaks and user preference index based on the LSTM optimization model;

[0081] S5: Obtain the real-time temperature of the magnetorheological fluid and the speed of the damper piston, establish a dynamic compensation model, and generate the final output current value;

[0082] S6: adjusting the damping force of the magnetorheological damper of the vehicle suspension system according to the final output current value;

[0083] The vehicle status data includes dimensionless lateral acceleration, pitch rate, and roll rate. Normalization can be used to eliminate dimensional differences and provide standardized input for stability coefficient calculation.

[0084] Road condition data includes dimensionless road roughness, tire slip, moisture content, and texture characteristic index. These data can be collected by sensors and processed through linear transformation to fully characterize complex road disturbance characteristics.

[0085] The fuzzy rule base is used to handle the nonlinear relationship between the vehicle stability coefficient and the road condition coefficient. Specifically, a Gaussian membership function can be used to define the fuzzy set of input variables, and the rule activation intensity is matched to the preset damping force.

[0086] The LSTM optimization model is used to learn the temporal correlation between the vehicle's historical vibration peaks and the user preference index. Specifically, the network weights can be trained by intercepting historical data in a time window to dynamically modify the basic damping force.

[0087] The dynamic compensation model is used to eliminate the interference of temperature and speed on the current output. Specifically, the linear compensation coefficient can be used to adjust and optimize the current value to ensure the accuracy of the final output current value.

[0088] Specifically, the vehicle's lateral acceleration, pitch rate, and roll rate are dimensionlessly processed and weighted to generate the vehicle's stability coefficient, reflecting the vehicle's dynamic balance. Road condition parameters such as road roughness and tire slip are dimensionlessly processed and weighted to generate a road condition coefficient, which characterizes the intensity of the road's comprehensive disturbance. A fuzzy rule base calculates the rule activation strength based on the membership of the stability coefficient and the road condition coefficient, and generates a basic damping force through weighted averaging, addressing the nonlinear mapping problem under multivariable coupling. An LSTM model analyzes historical vibration peak attenuation trends and user manual adjustment records to output an optimized damping force tailored to user preferences. Changes in magnetorheological fluid temperature are adjusted using a temperature compensation coefficient to adjust the current value, while piston speed is corrected using a speed compensation coefficient to eliminate environmental interference with the damping force.

[0089] Compared with existing technologies, traditional methods rely on a single vehicle state parameter threshold control and fail to integrate road moisture content or texture characteristic indices, resulting in inaccurate control on slippery roads. This solution, however, integrates four types of road condition parameters with three-axis vehicle state data, achieving multivariable collaborative decision-making through a fuzzy rule base. Existing technologies lack a user preference learning mechanism, while this solution incorporates an LSTM model to analyze manual adjustment history, enabling personalized adaptation of damping force output. Furthermore, existing methods fail to account for the impact of magnetorheological fluid temperature drift. This solution utilizes a dynamic compensation model to correct current output in real time, improving control stability under complex operating conditions.

[0090] Through the above technical solution, this application solves the problem of damping force inaccuracy caused by the lack of multi-dimensional road condition data. A fuzzy rule base is used to achieve a coordinated assessment of vehicle status and road disturbances, improving control accuracy on wet or rough roads. Simultaneously, the LSTM model learns user preferences and vibration attenuation patterns, dynamically adapting damping force output to individual needs. Temperature and speed compensation mechanisms eliminate the effects of material property drift and mechanical delays, ensuring the reliability of the final output current value in complex environments and achieving coordinated optimization of vehicle stability and ride comfort.

[0091] The present invention further proposes a method for generating the vehicle state stability coefficient, specifically:

[0092] By formula:

[0093] ;

[0094] Generate vehicle state stability coefficient K stab ;

[0095] In the formula, a y It represents the lateral acceleration of the vehicle, W pitch It represents the pitch angular rate, W roll It represents the roll angle rate, α, β, γ are all weight coefficients, and α+β+γ=1;

[0096] The vehicle's lateral acceleration refers to the acceleration of the vehicle in the lateral direction of motion. It can be measured using a three-axis acceleration sensor and normalized to obtain dimensionless data. It is used to reflect the centrifugal force when the vehicle turns.

[0097] Pitch rate refers to the angular velocity of the vehicle's rotation around its lateral axis. It can be measured using a gyroscope and processed through a low-pass filter to characterize the longitudinal undulation dynamics of the vehicle.

[0098] Roll rate refers to the angular velocity of the vehicle's rotation around its longitudinal axis. It can be collected using a six-axis inertial measurement unit and processed using a Kalman filter to reflect the lateral tilt trend of the vehicle body.

[0099] In addition, the weight coefficients α, β, and γ refer to the contribution ratio of each parameter in the stability assessment. Specifically, a dynamic adjustment algorithm can be used to update them in real time according to the driving scenario. For example, the α value can be increased when cornering at high speed to enhance lateral stability control.

[0100] Specifically, the vehicle's lateral acceleration, pitch rate, and roll rate are acquired through sensors and standardized to eliminate dimensional differences before being input into a weighted calculation model. Weight coefficients are dynamically assigned based on the real-time driving status. For example, when road bumps are detected, the pitch rate weight coefficient β is automatically increased to enhance longitudinal stability assessment. The linear combination of the three parameters covers the multi-dimensional dynamic characteristics of the vehicle during driving, and by constraining the sum of the weight coefficients to 1, it avoids a single parameter from excessively dominating the calculation results. Under braking conditions, the pitch rate weight β can be set to 0.5 to focus on suppressing the front end sinking phenomenon, while the roll rate weight γ is adjusted to 0.3 to take into account steering stability.

[0101] The present invention further proposes a method for generating the road condition coefficient, specifically:

[0102] By formula:

[0103] ;

[0104] Generate road condition coefficient K road ;

[0105] In the formula, R rough It represents the road roughness, s represents the tire slip rate, P represents the road moisture content, M tex It represents the texture feature index, k1, k2, k3, k4 are all weight coefficients, and k1+k2+k3+k4=1;

[0106] Road roughness refers to a quantitative indicator of the longitudinal undulation of the road surface during vehicle driving. It can be achieved by collecting elevation data using a lidar or inertial measurement unit and calculating the standard deviation to reflect the geometric characteristics of the macroscopic road surface.

[0107] Tire slip refers to the ratio of the tangential velocity difference between the tire and the road surface to the theoretical rolling speed. It can be calculated by the signal difference between the wheel speed sensor and the vehicle speed sensor, and is used to characterize the dynamic friction state between the tire and the road surface.

[0108] Pavement moisture content refers to the percentage of moisture in the pavement surface. It can be measured using a capacitive humidity sensor or an infrared reflective sensor to quantify the effect of water film on tire adhesion.

[0109] The texture characteristic index is a quantitative indicator of the microscopic roughness of the road surface. It can be calculated by extracting the eigenvalues ​​of the gray-level co-occurrence matrix of the road surface texture using image processing technology, and is used to describe the high-frequency vibration transmission characteristics.

[0110] In addition, the weight coefficient refers to the contribution ratio of each parameter to the road condition coefficient, which can be dynamically adjusted using the hierarchical analysis method or principal component analysis method to balance the priority of each parameter under different working conditions;

[0111] Specifically, the parameters of four dimensions, namely, road roughness, tire slip rate, road moisture content, and texture feature index, are normalized, and linear weighted fusion is used to generate the road condition coefficient after eliminating dimensional differences.

[0112] Among them, road roughness reflects the macroscopic undulations of the road surface, tire slip describes the dynamic contact between the tire and the road surface, road moisture content quantifies the effect of ambient humidity on friction, and the texture characteristic index describes the transmission characteristics of microscopic roughness to vibration. The contribution of each parameter is adjusted using dynamic weighting coefficients. For example, the weighting coefficient for road moisture content can be increased for slippery road conditions, while the weighting coefficient for texture characteristic index can be increased for rough road conditions. The resulting road condition coefficient comprehensively represents the road surface's geometric characteristics, friction properties, environmental interference, and vibration transmission characteristics, providing a unified and multi-dimensional input parameter for subsequent damping force control.

[0113] like Figure 2 As shown, the present invention further proposes a method for generating the basic damping force of the magnetorheological damper, which specifically includes the following steps:

[0114] S3.1: Based on the fuzzy rule base, generate the membership degree of the fuzzy rule set according to the vehicle state stability coefficient and the road condition coefficient;

[0115] S3.2: Generate rule activation strength based on the membership degree of the fuzzy rule set;

[0116] S3.3: Generate the basic damping force of the magnetorheological damper according to the rule activation intensity;

[0117] The membership of the fuzzy rule set refers to mapping the vehicle state stability coefficient and the road condition coefficient into the membership degree of the fuzzy set. Specifically, this can be achieved by using a Gaussian membership function. This function quantifies the fuzzification process of the input variables through the center value and width parameters, avoiding the rigid limitation of threshold segmentation.

[0118] Rule activation strength refers to the degree of matching between the vehicle state and road conditions to the current rule through a comprehensive evaluation of the membership product operation. Specifically, it can be generated by membership product, so that multiple fuzzy rules can be activated in parallel under different working conditions.

[0119] The basic damping force refers to the output value after weighted fusion of the preset damping force by activation intensity. Specifically, it can be generated by a weighted average algorithm, retaining the expert experience knowledge of the fuzzy rule base, and realizing a smooth transition of the damping force through real-time data driving.

[0120] Specifically, the vehicle state stability coefficient and road condition coefficient are first input into the fuzzy rule base, and the degree of membership of each in different fuzzy sets is calculated using a Gaussian membership function. For example, when the vehicle state stability coefficient is 0.8 and the road condition coefficient is 0.6, they correspond to the fuzzy sets of "high stability" and "medium road conditions," respectively, and the corresponding membership values ​​are calculated. Next, the activation strength of each fuzzy rule is generated through a membership product operation. For example, when a rule requires both "high stability" and "medium road conditions" to be met, its activation strength is the product of the two memberships. Finally, the preset damping forces corresponding to all activated rules are weighted and summed according to the activation strength, and then divided by the sum of the activation strengths to obtain the basic damping force output value. This achieves the dynamic fusion of vehicle state and road condition data, avoiding the control deviation caused by a single parameter or fixed threshold.

[0121] The present invention further proposes a method for generating the membership degree of the fuzzy rule set, specifically:

[0122] ;

[0123] Generate membership degree µ of fuzzy rule set A (x);

[0124] In the formula, x represents the input value, c represents the center value of the fuzzy set, and ω represents the set width; the input value x includes the vehicle state stability coefficient and the road condition coefficient;

[0125] The input value x is a combined input parameter of the vehicle state stability coefficient and the road condition coefficient. Specifically, it can be obtained by normalizing the vehicle lateral acceleration, pitch angular rate, roll angular rate, road surface roughness, and tire slip rate data collected by sensors to comprehensively reflect the vehicle dynamic characteristics and road surface characteristics.

[0126] The fuzzy set center value c refers to the core position parameter of the membership function, which can be determined through historical data training or dynamic adjustment algorithms to match the parameter distribution characteristics under different driving scenarios.

[0127] The set width ω refers to the distribution range parameter of the membership function, which can be adaptively adjusted according to the dynamic range of the input parameters to control the coverage area of ​​the fuzzy rules.

[0128] Specifically, the input value x is fuzzified by using the Gaussian membership function, and the vehicle state stability coefficient and road condition coefficient are used as joint input quantities to calculate their membership under each fuzzy rule;

[0129] The fuzzy set center value c and set width ω can be dynamically adjusted based on real-time data. For example, when a vehicle enters a slippery road, the weight of the road condition coefficient increases. At this time, by adjusting c and ω, the membership function shifts toward the high slip rate region, thereby enhancing the fuzzy rule's ability to match slippery roads.

[0130] During the membership calculation process, the fusion of the input value x enables the fuzzy rule base to simultaneously perceive changes in vehicle posture and road surface characteristics. For example, when the vehicle's roll rate suddenly increases and the road surface texture is rough, the coverage of the membership function is expanded by adjusting ω, ensuring that the rule activation strength can adapt to parameter fluctuations under complex operating conditions.

[0131] The present invention further proposes that the rule activation strength is generated in the following manner:

[0132] By formula:

[0133] ;

[0134] Generate rule activation strength gi;

[0135] In the formula, µ Kstab Represents the vehicle state stability coefficient K stab Membership degree at the current rule level, µ Kroad Indicates the road condition coefficient K road The degree of membership in the current rule level;

[0136] The rule activation strength refers to the degree to which the fuzzy rule is triggered under specific working conditions. It can be achieved through the membership product operation, which quantifies the correlation between the vehicle dynamic characteristics and the road surface characteristics into an activation strength value.

[0137] Membership refers to the degree of matching between the input parameters and the fuzzy set. It can be calculated using the Gaussian membership function. The center value and width are adjusted according to the actual working conditions to characterize the degree to which the parameters deviate from the ideal state.

[0138] The vehicle state stability coefficient membership reflects the stability deviation of the vehicle body motion state, while the road condition coefficient membership characterizes the intensity characteristics of the road surface excitation. The product operation of the two mathematically realizes the nonlinear relationship modeling of the vehicle-road coupling system.

[0139] Specifically, when the vehicle is in a specific driving condition, the Gaussian membership function is first used to calculate the membership of the vehicle state stability coefficient and the road condition coefficient at each fuzzy rule level. For example, when the vehicle's lateral acceleration increases, causing the stability coefficient to increase, its membership in the high-level rules increases accordingly. At the same time, if the road surface roughness increases, causing the road condition coefficient to increase, its membership in the high-level rules also increases synchronously. After multiplying the two memberships, the activation strength of the corresponding rule is generated. The larger the strength value, the higher the influence weight of the rule on the damping force control under the current working condition. By traversing all fuzzy rules and calculating the activation strength, the dominant rule set is finally screened out, providing a dynamic matching basis for the subsequent damping force calculation.

[0140] The present invention further proposes that the basic damping force of the magnetorheological damper is generated in the following manner:

[0141] By formula:

[0142] ;

[0143] Generate the basic damping force F of the magnetorheological damper base ;

[0144] In the formula, g i It represents the rule activation strength, F i It represents the preset damping force corresponding to the i-th rule, and n represents the total number of activated rules;

[0145] The rule activation strength refers to the applicability of each fuzzy rule under the current working conditions. It can be calculated by multiplying the vehicle state stability coefficient and the road condition coefficient by the membership degree. This parameter reflects the contribution weight of different rules in real-time control.

[0146] The preset damping force refers to the reference damping force value corresponding to the typical working condition preset in the fuzzy rule base. It can be obtained through expert experience or historical data training to provide a basic control benchmark.

[0147] The total number of activated rules refers to the number of fuzzy rules that currently meet the activation conditions. This can be achieved by setting a membership threshold to ensure that the rules involved in the calculation are relevant to the current working conditions.

[0148] Specifically, during vehicle driving, when the sensor detects parameters such as lateral acceleration and pitch angular rate, the system first calculates the membership of the vehicle state stability coefficient and the road condition coefficient, and then obtains the activation strength of each fuzzy rule; for example, when the vehicle is in a sharp turn, the rule activation strength corresponding to the high roll angular rate will be significantly improved, so that the preset damping force associated with this condition will obtain a greater weight; by weightedly summing the preset damping forces of each activated rule with their corresponding activation strengths, and then dividing them by the total activation strengths to complete the normalization process, the final output basic damping force not only retains the benchmark value set by expert experience, but also can dynamically adjust the weight ratio of each rule according to the real-time working conditions; this dynamic weighting mechanism effectively avoids the problem of a single rule dominating the control output in traditional methods.

[0149] like Figure 3 As shown, the present invention further proposes that the generation method of the optimized damping force of the magnetorheological damper specifically includes:

[0150] S4.1: Obtain vehicle historical vibration peak values ​​and user preference index;

[0151] S4.2: Establish an LSTM optimization model, substitute the vehicle's historical vibration peak value and user preference index into the LSTM model, and generate the optimized damping force of the magnetorheological damper;

[0152] The LSTM model is specifically:

[0153] ;

[0154] In the formula, LSTM stands for long short-term memory network, A peak It represents the historical vibration peak value of the vehicle, U pref It represents the user preference index, F base It represents the basic damping force of the magnetorheological damper;

[0155] The user preference index U pref The specific way to obtain it is:

[0156] By formula:

[0157] ;

[0158] In the formula, U adjust It represents the latest manual adjustment value of the user, U pref （t） It represents the user preference index, U pref （t-1） It represents the user preference index of the last update, and t represents the update number.

[0159] The vehicle's historical vibration peak value refers to the maximum vibration acceleration recorded by the vehicle's suspension system within a specific time period. This can be achieved by collecting data from an acceleration sensor and using a sliding window to calculate extreme values. It is used to reflect the vibration characteristic pattern of the vehicle during long-term driving.

[0160] The user preference index is a parameter that quantifies the user's personalized preferences for the softness or firmness of the suspension system. This index captures user manual adjustment records through the vehicle's human-machine interface and uses a recursive averaging algorithm to update historical data, balancing the stability of user habits with the sudden change in adjustment behavior.

[0161] The LSTM optimization model is a recurrent neural network model with long short-term memory units. It can be constructed using a deep learning framework that includes a forget gate, input gate, and output gate structure. It is used to capture dynamic features in time series data and predict the optimal value of the damping force.

[0162] Specifically, the vehicle's historical vibration peaks are collected by an accelerometer at a fixed sampling frequency, for example, every 10 milliseconds. After statistics are performed on a sliding window, the maximum vibration value in the past 5 seconds is extracted as the input feature.

[0163] During the update process of the user preference index, the weight coefficient in the recursive averaging formula can be dynamically adjusted according to the number of updates. For example, when n is less than 10 times, the latest adjustment value is given a higher weight to quickly respond to user needs. When n is greater than 50 times, the weight of the latest value is reduced to maintain preference stability.

[0164] When training the LSTM model, historical vibration peak values ​​and user preference indexes are combined into a time series input. The output layer is mapped to a damping force correction value through a fully connected network. The memory unit of the hidden layer retains state information within the past 30 seconds to identify vibration pattern change trends.

[0165] During the model inference process, the real-time collected vibration data and user adjustment behavior are encoded into time series feature vectors, which are forward-propagated through the LSTM network to generate optimized damping force parameters that adapt to the current driving state and user needs.

[0166] like Figure 4 As shown, the present invention further proposes that the final output current value is generated in the following manner:

[0167] S5.1: Generate an optimized current value based on the optimized damping force of the magnetorheological damper based on the damping force conversion formula;

[0168] S5.2: Obtain the real-time temperature of the magnetorheological fluid and the damper piston speed, establish a dynamic compensation model, and generate the final output current value;

[0169] The expression of the dynamic compensation model is specifically:

[0170] ;

[0171] In the expression, I final Indicates the final output current value, I optimal Indicates the optimized current value, K T It represents the temperature compensation coefficient of magnetorheological fluid, T j It represents the real-time temperature of the magnetorheological fluid, T ref It represents the calibration temperature of magnetorheological fluid, K V It represents the speed compensation coefficient, v piston It represents the damper piston speed;

[0172] The dynamic compensation model refers to a mathematical model that corrects the optimized current using the real-time collected magnetorheological fluid temperature and piston speed parameters. Specifically, the real-time data can be obtained by using temperature sensors and speed sensors, and a product relationship can be constructed in combination with preset compensation coefficients.

[0173] The temperature compensation coefficient is used to characterize the sensitivity of the viscosity of the magnetorheological fluid to temperature changes. It can be determined by fitting the current deviation at different temperatures in the laboratory.

[0174] The speed compensation coefficient is used to characterize the degree of influence of the piston movement speed on the magnetic field response delay, which can be obtained by testing the correlation between the piston speed and the damping force lag time;

[0175] The MR fluid calibration temperature refers to the temperature at which the viscosity of the MR fluid is at the design reference state, which is usually set to a constant value within the range of 20-25 degrees Celsius.

[0176] Specifically, after generating the optimized current value, the MR fluid's operating temperature is monitored in real time via a temperature sensor, while the piston's velocity is measured via a displacement sensor. When the temperature rises above the calibration temperature, the MR fluid's viscosity decreases, resulting in a lower damping force at the same current. The temperature compensation term then generates a positive current increment to offset this deviation. As the piston speed increases, the magnetic field's response delay leads to insufficient instantaneous damping force. The velocity compensation term compensates for this lag by increasing the current output. The dynamic compensation model multiplies the optimized current value by the two compensation terms, enabling adaptive adjustment of the final output current. For example, it automatically increases the current to maintain the target damping force under high-temperature conditions, or enhances the magnetic field strength in advance during high-speed piston movement.

[0177] like Figure 5 As shown, the present invention further proposes a control system for a magnetorheological damper of a vehicle suspension system, comprising a vehicle state analysis unit 10, a road condition analysis unit 20, a basic damping force analysis unit 30, a damping force optimization unit 40, a final output current value generation unit 50 and a control unit 60;

[0178] The vehicle state analysis unit collects the vehicle's lateral acceleration, pitch rate, and roll rate through onboard sensors and performs dimensionless processing on the multi-source data. Specifically, a normalization algorithm is used to eliminate dimensional differences and address the interference of different physical quantities on stability assessment.

[0179] The road condition analysis unit is a multi-dimensional detection module that integrates road surface roughness, tire slip rate, road surface moisture content, and texture feature index. Specifically, it uses LiDAR and image sensor fusion technology to enhance the ability to characterize complex road surface features.

[0180] The basic damping force analysis unit is a decision-making module that performs nonlinear mapping between vehicle stability coefficient and road condition coefficient based on a fuzzy rule base. Specifically, it can use Gaussian membership functions to construct fuzzy rules, breaking through the rigid constraints of traditional threshold control.

[0181] The damping force optimization unit is a learning module that uses an LSTM model to extract historical vibration peak time series features and combines them with the user preference index. Specifically, it uses a sliding time window to store historical data to achieve progressive learning of personalized driving habits.

[0182] The final output current value generation unit refers to a correction module that dynamically compensates for the magnetorheological fluid temperature and piston speed. Specifically, a temperature sensor and a speed sensor can be used to collect environmental parameters in real time to solve the current output offset problem.

[0183] The control unit refers to the output module that converts the final output current value into the actuator control signal. Specifically, a PID controller can be used to achieve closed-loop regulation;

[0184] Specifically, the vehicle state analysis unit normalizes lateral acceleration, pitch rate, and roll rate to generate a dimensionless vehicle stability coefficient, eliminating interference from parameters of varying dimensions on the evaluation model. The road condition analysis unit uses multi-sensor fusion to obtain road surface roughness, slip ratio, moisture content, and texture feature index. After dimensionless processing, it generates a comprehensive road condition coefficient, improving data representation accuracy under complex road conditions. The basic damping force analysis unit uses a fuzzy rule base to perform a nonlinear mapping between the vehicle stability coefficient and the road condition coefficient. It calculates the rule activation strength using a Gaussian membership function to generate a basic damping force adapted to different operating conditions. The damping force optimization unit utilizes an LSTM model to extract the temporal characteristics of historical vibration peaks. In combination with a user preference index update mechanism, it transforms manual adjustments into an optimization strategy, generating an optimized damping force that balances both objective physical conditions and subjective driving habits. The final output current generation unit uses a dynamic compensation model to adjust the magnetorheological fluid temperature and piston speed in real time, adjusting the optimized current value to generate the final output current, ensuring control accuracy under fluctuating environmental parameters. The control unit transmits the final output current to the magnetorheological damper actuator, forming a closed-loop control link from data acquisition to damping force adjustment.

[0185] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for controlling a magnetorheological damper for a vehicle suspension system, characterized in that: The following steps are involved: Acquiring vehicle state data and generating a vehicle state stability coefficient; wherein the vehicle state data includes vehicle lateral acceleration, pitch angular rate, and roll angular rate; Obtaining road condition data of the road on which the vehicle is traveling and generating a road condition coefficient; wherein the road condition data includes road surface roughness, tire slip rate, road surface moisture content, and texture characteristic index; the texture characteristic index refers to the roughness of the road surface texture; Based on the fuzzy rule base, the basic damping force of the magnetorheological damper is generated according to the vehicle state stability coefficient and the road condition coefficient; Based on the LSTM optimization model, the optimized damping force of the magnetorheological damper is generated according to the vehicle's historical vibration peak value and user preference index; Obtain the real-time temperature of the magnetorheological fluid and the speed of the damper piston, establish a dynamic compensation model, and generate the final output current value; Adjusting the damping force of the magnetorheological damper of the vehicle suspension system according to the final output current value; The method for generating the basic damping force of the magnetorheological damper specifically includes the following steps: Based on the fuzzy rule base, the membership degree of the fuzzy rule set is generated according to the vehicle state stability coefficient and the road condition coefficient; Generate rule activation strength based on the membership degree of the fuzzy rule set; Generate the basic damping force of the magnetorheological damper according to the rule activation intensity; The rule activation strength is generated in the following manner: By formula: ； Generate rule activation strength g i ; In the formula, µ Kstab Represents the vehicle state stability coefficient K stab Membership degree at the current rule level, µ Kroad Indicates the road condition coefficient K road The degree of membership in the current rule hierarchy; The basic damping force of the magnetorheological damper is generated in the following manner: By formula: ； Generate the basic damping force F of the magnetorheological damper base ; In the formula, g i It represents the rule activation strength, F i It represents the preset damping force corresponding to the i-th rule, and n represents the total number of activated rules.

2. The method for controlling a magnetorheological damper for a vehicle suspension system according to claim 1, wherein: The vehicle state stability coefficient is generated in the following manner: By formula: ； Generate vehicle state stability coefficient K stab ; In the formula, a y It represents the lateral acceleration of the vehicle, W pitch It represents the pitch angular rate, W roll It represents the roll angle rate, α, β, and γ are all weight coefficients, and α+β+γ=1.

3. The method for controlling a magnetorheological damper for a vehicle suspension system according to claim 1, wherein: The road condition coefficient is generated in the following manner: By formula: ； Generate road condition coefficient K road ; In the formula, R rough It represents the road roughness, s represents the tire slip rate, P represents the road moisture content, M tex It represents the texture feature index, k1, k2, k3, k4 are all weight coefficients, and k1+k2+k3+k4=1.

4. The method for controlling a magnetorheological damper for a vehicle suspension system according to claim 1, wherein: The membership degree of the fuzzy rule set is generated in the following manner: ； Generate membership degree µ of fuzzy rule set A (x); In the formula, x represents the input value, c represents the center value of the fuzzy set, and ω represents the set width; the input value x includes the vehicle state stability coefficient and the road condition coefficient.

5. The method for controlling a magnetorheological damper for a vehicle suspension system according to claim 1, wherein: The method for generating the optimized damping force of the magnetorheological damper specifically includes: Obtain vehicle historical vibration peak values ​​and user preference index; An LSTM optimization model was established, and the vehicle's historical vibration peak value and user preference index were substituted into the LSTM model to generate the optimized damping force of the magnetorheological damper. The LSTM model is specifically: ； In the formula, LSTM stands for long short-term memory network, A peak It represents the historical vibration peak value of the vehicle, U pref It represents the user preference index, F base It represents the basic damping force of the magnetorheological damper; The user preference index U pref The specific way to obtain it is: By formula: ； In the formula, U adjust It represents the latest manual adjustment value of the user, U pref （t） It represents the user preference index, U pref （t -1） It represents the user preference index of the last update, and t represents the update number.

6. The method for controlling a magnetorheological damper for a vehicle suspension system according to claim 1, wherein: The final output current value is generated in the following manner: Based on the damping force conversion formula, the optimized current value is generated according to the optimized damping force of the magnetorheological damper; Obtain the real-time temperature of the magnetorheological fluid and the speed of the damper piston, establish a dynamic compensation model, and generate the final output current value; The expression of the dynamic compensation model is specifically: ； In the expression, I final Indicates the final output current value, I optimal Indicates the optimized current value, K T It represents the temperature compensation coefficient of magnetorheological fluid, T j It represents the real-time temperature of magnetorheological fluid, T ref It represents the calibration temperature of magnetorheological fluid, K V It represents the speed compensation coefficient, v piston It represents the damper piston speed.

7. A control system for a magnetorheological damper of a vehicle suspension system, characterized in that: The system is used to execute a control method for a magnetorheological damper for a vehicle suspension system according to any one of claims 1 to 6, specifically comprising: A vehicle state analysis unit is used to obtain vehicle state data and generate a vehicle state stability coefficient; wherein the vehicle state data includes the vehicle's lateral acceleration, pitch angular rate, and roll angular rate; the vehicle's lateral acceleration, pitch angular rate, and roll angular rate are all dimensionless data; A road condition analysis unit is used to obtain road condition data of the road on which the vehicle is traveling and generate a road condition coefficient; the road condition data includes road surface roughness, tire slip rate, road surface moisture content, and texture characteristic index; the texture characteristic index refers to the roughness of the road surface texture; the road surface roughness, tire slip rate, road surface moisture content, and texture characteristic index are all dimensionless data; A basic damping force analysis unit is used to generate the basic damping force of the magnetorheological damper based on the fuzzy rule base, the vehicle state stability coefficient and the road condition coefficient; A damping force optimization unit, which generates the optimized damping force of the magnetorheological damper based on the LSTM optimization model, the vehicle's historical vibration peak value, and the user preference index; A final output current value generating unit is used to obtain the real-time temperature of the magnetorheological fluid and the speed of the damper piston to generate a final output current value; The control unit is used to adjust the damping force of the magnetorheological damper of the vehicle suspension system according to the final output current value.

Images