Motor Control and Efficiency Maps
An EV motor controller turns the driver's torque request into precisely timed phase currents while respecting the limits of the battery, inverter, motor, tyres, and cooling system. The familiar accelerator response of an EV is therefore a real-time control result, not an inherent motor characteristic.
The control chain
The accelerator pedal does not command a motor voltage directly. Vehicle software interprets pedal position alongside drive mode, wheel slip, stability control, battery state, thermal limits, and component faults. It then requests positive or negative axle torque.
A typical traction controller uses nested loops. A slower speed or torque loop determines the required motor current, while faster current loops regulate the electromagnetic state many thousands of times per second. The controller measures phase current and DC-link voltage, estimates or measures rotor position, calculates target phase voltages, and asks the traction inverter to reproduce them through pulse-width modulation.
The result must remain stable while speed, torque, battery voltage, temperature, and magnetic saturation change. Control quality affects efficiency, acoustic behavior, regenerative braking, traction, and component stress.
Field-oriented control
Three-phase stator currents are separated by 120 electrical degrees, which makes direct control cumbersome. Field-oriented control, also called vector control, mathematically transforms those measured currents into a coordinate system that rotates with the motor's magnetic field.
The transformed current has two principal components:
- d-axis current primarily changes magnetic flux;
- q-axis current primarily produces torque in a simple surface-magnet machine.
The distinction is not absolute for every topology. In an interior permanent-magnet motor, the interaction between d- and q-axis current also produces reluctance torque. Induction motors need an estimate of rotor flux and slip. Electrically excited machines add rotor-field current as another control variable.
After the controller calculates the desired d- and q-axis voltages, inverse transforms convert them back into stationary phase commands. Space-vector or another PWM method then selects inverter switching states whose average voltage follows the requested vector.
Rotor position and speed
Accurate vector control needs the electrical angle of the rotor field. A resolver, magnetic encoder, inductive sensor, or other position sensor can provide this angle directly. The controller combines position samples over time to obtain speed.
Sensorless control instead estimates angle and speed from measured current, commanded voltage, back electromotive force, inductance variation, or an observer model. This can reduce hardware and wiring, but estimation is difficult near zero speed because back-EMF information becomes weak. Some drives combine sensor data and model-based estimation for plausibility checking or degraded operation.
An angular error rotates the controller's assumed d-q frame away from the real rotor. That can reduce available torque, increase current and heat, create ripple, or threaten stability. Position sensing is therefore a performance and functional-safety component, not merely a speedometer.
MTPA, base speed, and field weakening
At low and medium speed, the controller usually seeks the required torque with the least useful stator current. Maximum torque per ampere, or MTPA, selects a d-q current combination that minimizes current for a given torque. For a surface-magnet motor this may be close to zero d-axis current, while a salient interior-magnet motor can benefit from negative d-axis current that extracts reluctance torque.
As motor speed rises, back EMF and inductive voltage consume more of the voltage available from the inverter. Base speed is the approximate point where voltage, rather than current, becomes the active boundary. Above it, the controller uses field weakening.
In a permanent-magnet motor, negative d-axis current creates a stator field that opposes part of the magnet flux linking the stator. It does not turn down the permanent magnets themselves. The lower net flux linkage restrains back EMF and permits higher speed, but current capacity is then shared between flux weakening and torque production. Torque falls and additional copper loss appears.
At still higher speed, maximum torque per volt, or MTPV, can choose the current angle that makes the best use of the remaining voltage. Current, voltage, rotor-speed, demagnetization, and thermal boundaries together create the familiar constant-torque and approximately constant-power regions.
Efficiency maps and optimal control
The current command that minimizes electrical current is not always the command that minimizes total drive-unit loss. Copper loss, iron loss, magnet loss, inverter switching and conduction loss, rotor excitation loss, gear loss, and cooling-pump power respond differently to speed and torque.
Production controllers therefore use calibrated maps and physical models. At each operating point they may choose:
- d- and q-axis current targets;
- rotor excitation in an electrically excited motor;
- PWM strategy and switching frequency;
- front-versus-rear torque in a multi-motor vehicle;
- whether an axle should be de-energized or disconnected;
- allowable torque based on present and predicted temperatures.
Efficiency optimization must coexist with response, noise, voltage ripple, battery limits, and reserve margins. A controller that chases a laboratory optimum too aggressively may create poor drivability or insufficient fault tolerance.
Regeneration and torque accuracy
Regenerative braking reverses power flow but uses the same current-control foundation. The vehicle requests negative wheel torque, the motor generates electrical power, and the inverter returns controlled DC current to the high-voltage bus.
Available regeneration may be restricted by battery state of charge, cell temperature, charge-power limits, tyre grip, motor speed, or inverter temperature. Brake blending fills the difference with friction brakes while maintaining the driver's requested deceleration and vehicle stability.
Torque accuracy also matters during positive drive. Stability and traction systems rely on the drive unit producing the commanded torque quickly and predictably. The controller may estimate actual torque from current, rotor position, calibrated motor parameters, and temperature because a production vehicle rarely carries a torque sensor on each motor shaft.
Fault handling
The controller continuously checks phase currents, DC-link voltage, rotor position, semiconductor temperature, motor temperature, isolation status, and communications. Plausibility checks compare redundant information. If measurements disagree or a limit is crossed, the system may reduce torque, disable one motor, stop switching, or open high-voltage contactors.
Turning off semiconductor gates does not make every motor electrically inert. A spinning permanent-magnet motor can continue generating voltage. Safe-state design therefore considers open-circuit back EMF, short-circuit torque, overspeed, inverter failure modes, and whether another axle can maintain controlled propulsion.
Continue through the motor series
Return to Electric Motors and Drive Units for the full motor and drive-unit series.