Road vehicle noise plays a large part in whether a vehicle commercially succeeds. We can greatly reduce both engine and road noise in a car cabin with passive acoustic treatments such as structural damping and acoustic absorption. However, recent interest in active noise control has been driven by the need to improve vehicle fuel efficiency through new economical engine designs and reduced vehicle weight.
Economical engine designs (such as variable displacement, which usually operates by deactivating cylinders) often result in increased low-frequency noise, as does reducing vehicle weight. Low-frequency noise is difficult to control using lightweight passive measures, and because ANC systems are most effective at low frequencies and result in a relatively small increase in weight, they offer a convenient complementary solution.
The main factor limiting the widespread commercial implementation of ANC systems in vehicles is the high implementation cost. Therefore, we must integrate the control system into the standard car electronic system. We focused on developing a cost-effective ANC system, with the potential for widespread commercial implementation.
ANC Strategies in Road Vehicles
Feedforward engine noise control is one example of an ANC that has been thoroughly investigated and implemented by a number of major car manufacturers. These systems use the car audio speakers to cancel the engine noise in the car cabin, which is monitored using an array of error microphones. A reference signal obtained from the engine provides engine speed and drives the speakers via an adaptive filter to minimise pressure at the microphones. By integrating with the audio system, this controller requires very little additional hardware or expense.
Road noise reduction using a feedforward control system must obtain multiple reference signals from accelerometers mounted on the vehicular structure. This results in an expensive system that is difficult to implement.
A feedback controller, which does not require a reference signal, is well-suited to control road noise. Standard single-input, single-output feedback control systems are successfully used in noise-reducing headphones, but they can achieve only limited control throughout the car cabin. Therefore, we developed a feedback control system using error sensors and speakers. The system uses the error microphones in the feedforward engine noise control system and the car audio speakers and therefore requires little additional hardware. This road noise control system architecture keeps implementation costs low.
Measuring the Vehicle Acoustic Environment
To predict the performance of noise control strategies in a small car, we measured the acoustic response between the eight error microphones and the four car audio speakers. Additionally, we measured the acoustic disturbance produced by both engine and road noise at the eight error microphones. To facilitate these measurements, the data acquisition system gave us at least 17 channels of simultaneously sampled input with antialiasing filters, and at least one channel of audio output.
Additionally, due to road testing, we needed to use the car’s 12 V power supply to power the system. We fulfilled these requirements with an NI CompactDAQ system with five NI 9234 C Series dynamic signal acquisition modules and an NI 9263 C Series analogue output module. The compatibility of NI CompactDAQ with a widely used, text-based mathematics language significantly benefitted our application. We could quickly obtain results because we had experience with this software environment, and the required code was already written.
The resulting engine and road noise measurements highlighted the frequency regions requiring ANC. For example, the engine noise measurements showed a strong narrowband peak in the spectrum at the frequency of the second engine order produced by the tested car’s four-cylinder engine. Conversely, the road noise measurements highlighted a broadband peak between 80 Hz and 160 Hz, which is the component targeted for feedback control.
Predicting the Performance of the ANC System
We can assess the performance of the ANC strategies from the reduction in the acoustic potential energy. Using the measured data, we synthesized the performance of the feedforward controller offline and produced a reduction in the noise of the second engine on the predicted order of between 5 dB and 12 dB. We predicted the performance of the feedback controller using the measured data and a reduction across the targeted frequency range of around 6 dB.
We used NI CompactDAQ to create an active engine and road noise control system for integrated implementation in a car. Using NI CompactDAQ I/O capabilities, simultaneous sampling, antialiasing filters, and DC powering, we can accurately measure the acoustic environment in a car cabin. We used the measured data to predict controller performance in reducing both engine and road noise, and these predictions indicate a useful level of low-frequency noise reduction. Future work will involve a real-time implementation of these systems. We expect significant cost savings when we update our system because we can use the same input and output modules with NI CompactDAQ.
Institute of Sound and Vibration Research, University of Southampton
Southampton SO17 1BJ