How do you implement a sliding mode observer for sensorless control of a brushless DC motor (BLDC)?

1 What is a SMO?

A SMO is a type of observer that uses a sliding mode control (SMC) technique to estimate the state variables of a nonlinear system. SMC is a feedback control method that forces the system to operate on a predefined sliding surface, where the error between the actual and desired states is zero. The sliding surface is designed to ensure stability, performance, and robustness of the system. The SMO uses the same sliding surface as the SMC, but instead of controlling the system input, it estimates the system output based on the system input and a switching function. The switching function is designed to drive the estimation error to zero in finite time, regardless of the system uncertainties and disturbances.

2 How to design a SMO for BLDC?

To design a SMO for BLDC, you need to first model the BLDC motor as a nonlinear system with three inputs (the phase currents) and three outputs (the back electromotive forces). The state variables of the system are the rotor position and speed, which are not directly measurable. The SMO will estimate these state variables based on the inputs and outputs of the system. The sliding surface for the SMO is defined as the difference between the actual and estimated back electromotive forces, which should be zero when the estimation is accurate. The switching function for the SMO is defined as a sign function of the sliding surface, which switches between positive and negative values depending on the sign of the error. The SMO then updates the state estimates using a linear combination of the inputs, outputs, switching function, and some tuning parameters.

3 How to implement a SMO for BLDC?

To implement a SMO for BLDC, you need to program the SMO algorithm in a microcontroller or a digital signal processor (DSP) that can communicate with the motor driver and the current sensors. The SMO algorithm requires sampling the phase currents and calculating the back electromotive forces at a high frequency, typically higher than the switching frequency of the motor driver. The SMO algorithm also requires tuning the parameters of the sliding surface and the state update equations, such as the gains, the bounds, and the initial conditions. These parameters affect the speed and accuracy of the estimation, as well as the robustness and stability of the observer. You can use simulation tools or experimental methods to find the optimal values of these parameters for your specific application.

4 What are the advantages of SMO for BLDC?

The main advantage of SMO for BLDC is that it can provide accurate and robust estimation of the rotor position and speed, even under variable load and speed conditions. This can improve the performance and efficiency of the motor drive system, and enable sensorless control strategies that require precise knowledge of the rotor position and speed, such as vector control or field-oriented control. Another advantage of SMO for BLDC is that it is relatively simple and easy to implement, compared to other sensorless methods that require complex signal processing or model identification techniques. The SMO only requires measuring the phase currents and calculating the back electromotive forces, which are readily available in most motor drive systems.

5 What are the disadvantages of SMO for BLDC?

The main disadvantage of SMO for BLDC is that it can cause high-frequency chattering in the estimation, due to the switching function that changes rapidly between positive and negative values. This can introduce noise and harmonic distortion in the system, and affect the quality and reliability of the estimation. To reduce the chattering, you can use some techniques such as boundary layer, hysteresis, or adaptive switching, but these may compromise the speed and accuracy of the estimation, or increase the complexity of the algorithm. Another disadvantage of SMO for BLDC is that it may have difficulties in estimating the initial rotor position and speed, especially at low or zero speed. This can cause errors or delays in starting or reversing the motor, or in detecting faults or changes in the load. To overcome this problem, you can use some methods such as injecting high-frequency signals, applying initial pulses, or combining with other sensorless methods.

6 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?