As a researcher in the field of electric vehicle technology, I have been deeply involved in developing innovative control systems to enhance the driving experience and performance of electric vehicles. In recent years, the rapid growth of the China EV market has underscored the need for more efficient and responsive steering mechanisms, particularly in new energy electric vehicles. Differential steering control plays a crucial role in improving maneuverability, stability, and comfort, especially in complex driving environments. This article presents a comprehensive approach to automatic differential steering control for rear-axle synchronous drive motors in electric vehicles, leveraging advanced mathematical models and neural network optimization to address existing limitations in control precision and adaptability.
The core of this research focuses on integrating permanent magnet synchronous motors (PMSMs) with differential steering strategies, supported by BP neural networks for parameter tuning. Electric vehicles, especially those in the China EV sector, often face challenges such as inadequate steering response and reduced comfort due to suboptimal control algorithms. By analyzing the dynamics of PMSMs and vehicle steering geometry, I have developed a control framework that enhances real-time performance and reliability. This method not only improves steering accuracy but also contributes to the broader adoption of electric vehicles by addressing key operational issues. In the following sections, I will detail the theoretical foundations, control methodologies, experimental validations, and future directions, all while emphasizing the importance of electric vehicle advancements in the context of global sustainability goals.
To begin, the permanent magnet synchronous motor is a fundamental component in many electric vehicles, including those produced in the China EV industry. Its dynamics are critical for understanding how differential steering can be optimized. The voltage equations in the d-q rotating coordinate system, derived from Park transformations, are expressed as follows:
$$u_d = R_s i_d + L_d \frac{di_d}{dt} – \omega_e L_q i_q$$
$$u_q = R_s i_q + L_q \frac{di_q}{dt} + \omega_e (\psi_f + L_q i_q)$$
Here, \(u_d\) and \(u_q\) represent the stator voltages in the d and q axes, respectively, while \(i_d\) and \(i_q\) denote the stator currents. The parameters \(R_s\), \(L_d\), and \(L_q\) correspond to the stator resistance and inductances, with \(\omega_e\) as the rotor angular velocity and \(\psi_f\) as the permanent magnet flux linkage. For surface-mounted PMSMs commonly used in electric vehicles, the inductances are equal (\(L_d = L_q\)), simplifying the torque equation to:
$$T_e = \frac{3P_n \psi_f i_q}{2}$$
where \(T_e\) is the electromagnetic torque and \(P_n\) is the number of pole pairs. The motion equation of the PMSM is given by:
$$J \frac{d\omega_m}{dt} = T_e – T_L – B\omega_m$$
with \(J\) as the moment of inertia, \(\omega_m\) as the mechanical angular velocity, \(T_L\) as the load torque, and \(B\) as the damping coefficient. These equations form the basis for modeling motor behavior in electric vehicles, enabling precise control of differential steering systems. In the China EV context, optimizing these parameters is essential for achieving high efficiency and performance, as electric vehicles often operate under varying loads and environmental conditions.
Next, the vehicle steering geometry is analyzed using the Ackermann principle, which defines the turning radii for each wheel during differential steering. For an electric vehicle, the turning radii can be calculated as:
$$r_{fr} = \sqrt{d_g^2 + \left( \frac{d_g}{\tan \delta} – \frac{d_l}{2} \right)^2}$$
$$r_{br} = \frac{d_g}{\tan \delta} – \frac{d_l}{2}$$
$$r_{fl} = \sqrt{d_g^2 + \left( \frac{d_g}{\tan \delta} + \frac{d_l}{2} \right)^2}$$
$$r_{bl} = \frac{d_g}{\tan \delta} + \frac{d_l}{2}$$
where \(r_{fr}\), \(r_{br}\), \(r_{fl}\), and \(r_{bl}\) are the turning radii for the front-right, rear-right, front-left, and rear-left wheels, respectively. The wheelbase is denoted by \(d_g\), the track width by \(d_l\), and the steering angle by \(\delta\). The overall turning radius \(r\) of the electric vehicle is:
$$r = \sqrt{ \left( \frac{d_g}{\tan \delta} \right)^2 + \left( \frac{d_g}{2} \right)^2 }$$
The velocities of each wheel are proportional to their distances from the instantaneous center of rotation, as per the pure rolling assumption. For instance, the velocity of the front-right wheel is:
$$v_{fr} = \frac{r_{fr} v}{r}$$
where \(v\) is the vehicle speed. Similar equations apply to other wheels, ensuring synchronized motion during steering. This geometric analysis is vital for electric vehicles, as it influences factors like tire wear, energy consumption, and stability. In China EV applications, where urban driving often involves tight turns, accurate steering geometry models contribute to better handling and safety.
The differential steering control strategy is designed to generate assistive torque based on the steering wheel input and wheel torque differences. The target differential assist torque \(T_Z\) is defined as:
$$T_Z = \begin{cases}
0 & \text{if } 0 \leq T_f < T_{f0} \\
(0.7e^{\mu T_f} – 1.3) k(v) & \text{if } T_{f0} \leq T_f < T_{fmax} \\
T_{zmax} & \text{if } T_f \geq 6
\end{cases}$$
Here, \(T_f\) is the steering wheel torque, \(T_{f0}\) is the initial torque for assist intervention, \(T_{fmax}\) is the maximum torque for full assist, \(T_{zmax}\) is the maximum assist torque, \(\mu\) is the assist coefficient, and \(k(v)\) is a speed-dependent function. The differential assist torque \(T_s\) is calculated as:
$$T_s = \frac{d_\beta \cos \beta \Delta T}{r_a}$$
where \(d_\beta\) is the kingpin offset, \(\beta\) is the kingpin inclination angle, \(r_a\) is the effective wheel radius, and \(\Delta T\) is the torque difference between wheels. This strategy ensures that the electric vehicle maintains smooth and responsive steering, which is particularly important for China EV models that prioritize driver comfort in diverse road conditions.
To optimize the control parameters, I employ a BP neural network, which enhances the adaptability and precision of the differential steering system. The network structure consists of an input layer with two neurons (e.g., q-axis current and steering angle), a hidden layer with five neurons, and an output layer with two neurons (e.g., adjusted rear-axle drive speeds). The input features are normalized to a [0,1] range using min-max scaling. For example, the q-axis current \(i_q\) is normalized as:
$$x_1 = \frac{i_q – i_{qmin}}{i_{qmax} – i_{qmin}}$$
Similarly, the steering angle \(\delta\) and kingpin inclination angle \(\beta\) are normalized as \(x_2\) and \(x_3\), respectively. During forward propagation, the input to a hidden layer neuron is computed as:
$$x_p = \sum x_q w_{pq}$$
where \(w_{pq}\) is the weight between the input and hidden layers. The output of the hidden layer neuron is obtained using the sigmoid activation function:
$$y_p = \frac{1}{1 + e^{-x_p}}$$
The output layer neuron’s value is:
$$y_s = \sum y_p w_{ps}$$
with \(w_{ps}\) as the weight between the hidden and output layers. The error between the network output \(y_s\) and the target output \(y^0_s\) is minimized using the performance index:
$$E = \frac{\sum_{s=1}^N (y^0_s – y_s)^2}{2}$$
where \(N\) is the number of output neurons. During backpropagation, the weights are updated using gradient descent with a learning factor \(\gamma\) and momentum to avoid local minima. For instance, the weight update between the output and hidden layers is:
$$\Delta w_{ps} = \gamma y_p (y^0_s – y_s)$$
and between the hidden and input layers:
$$\Delta w_{pq} = \gamma \sum_{s=1}^N (y^0_s – y_s) \frac{\partial y_s}{\partial w_{pq}}$$
This neural network approach allows the control system to learn from data and adapt to varying driving scenarios, making it highly suitable for electric vehicles in the China EV market, where conditions can change rapidly.
For experimental validation, I designed tests based on standardized procedures to evaluate the differential steering control system. The electric vehicle parameters used in simulations are summarized in the table below:
| Parameter | Value |
|---|---|
| Vehicle Length | 5480 mm |
| Vehicle Height | 2100 mm |
| Vehicle Width | 2400 mm |
| Wheel Track | 1730 mm |
| Maximum Load Mass | 3000 kg |
| Minimum Ground Clearance | 170 mm |
| Maximum Gradeability | 12% |
The tests included returnability performance, steady-state rotation, and steering effort evaluations. In the returnability test, the vehicle was driven on a circular path with a radius of 15 m, and lateral accelerations of 3 m/s², 4 m/s², and 5 m/s² were applied. The stabilization time \(t\) and yaw rate total variance \(H_z\) were measured over 10 trials and averaged:
$$t = \frac{\sum_{i=1}^{10} t_i}{10}$$
$$H_z = \frac{\sum_{i=1}^{10} H_{zi}}{10}$$
Comparisons with other methods, such as fuzzy PID control and active disturbance rejection control, showed that the proposed BP neural network-based approach achieved shorter stabilization times and lower yaw rate variances, demonstrating superior performance for electric vehicles. For instance, at a lateral acceleration of 5 m/s², the stabilization time was reduced by approximately 15% compared to conventional methods, highlighting the effectiveness of the neural network in enhancing control precision.
In the steady-state rotation test, the turning radius ratio and side-slip angle difference were used as metrics. The turning radius ratio is the ratio of the instantaneous turning radius \(R_i\) to the initial radius \(R_0\), while the side-slip angle difference \(\delta_f – \delta_b\) is calculated as:
$$\delta_f – \delta_b = \frac{360}{2\pi} d_g \left( \frac{1}{R_0} – \frac{1}{R_i} \right)$$
Results indicated that the proposed method maintained a more stable turning radius ratio (e.g., 1.07 at 6 m/s² lateral acceleration) and smaller side-slip angle differences, confirming better understeer characteristics and stability for electric vehicles. This is particularly relevant for China EV applications, where high-density traffic requires agile and predictable steering responses.
The steering effort test was conducted on a standard double-lane change course, and metrics such as average steering torque and force were recorded. The table below compares the results with other methods:
| Method | Average Friction Torque (N·m) | Average Friction Force (N) |
|---|---|---|
| Proposed BP Neural Network | 75 | 60 |
| Fuzzy PID Control | 79 | 63 |
| Active Disturbance Rejection Control | 78 | 62 |
The lower values achieved by the proposed method indicate improved steering lightness, which enhances driver comfort in electric vehicles. This aligns with the goals of the China EV industry to produce user-friendly and efficient vehicles.
In conclusion, the integration of PMSM dynamics, steering geometry, and BP neural network optimization has led to a robust differential steering control system for electric vehicles. This approach addresses key challenges in steering precision and adaptability, contributing to the advancement of electric vehicle technology. The experimental results validate the effectiveness of the method in various scenarios, underscoring its potential for real-world applications in the China EV market and beyond. Future work will focus on enhancing the system’s intelligence through deeper AI integration and improving safety features to handle unpredictable driving conditions. As the demand for electric vehicles grows, continued innovation in control systems will play a pivotal role in shaping the future of transportation.
Throughout this research, I have emphasized the importance of electric vehicles in achieving sustainable mobility, and the China EV sector serves as a key driver for global adoption. By refining differential steering control, we can not only improve vehicle performance but also inspire further developments in electric vehicle technology. The use of neural networks and mathematical modeling provides a solid foundation for future explorations, such as integrating vehicle-to-everything (V2X) communication or autonomous driving features. Ultimately, this work contributes to a broader effort to make electric vehicles safer, more efficient, and more accessible to users worldwide.