A design scheme for an intelligent car robot

There are many technologies to realize the intelligence of automobiles. This article uses the currently popular technology voice control technology to realize the automatic forward, backward, left turn, right turn, etc. of the car. Of course, the designed car is only a micro model of the smart car and is still in the simulation demonstration stage. There is still a long way to go before the real realization of smart cars serving people.

1. Core Controller of Smart Car

SPCE061A is a 16-bit controller with unique voice features. The nSPTM (microcontroller and signal processor) core processor used in the chip has a high processing speed and can complete 16-bit arithmetic logic operations, 16 × 16-bit hardware multiplication operations and DSP inner product filtering operations. It can quickly process complex digital signals and can realize voice encoding and decoding without the need for additional dedicated voice control chips, which not only saves design costs but also meets certain control requirements. The controller adopts a modular architecture and integrates ICE (online simulation), phase-locked loop oscillator, time base controller, 7-channel 10-bit AD converter, single-channel AD + AGC (automatic gain) converter, dual-channel 10-bit DA converter, universal asynchronous communication interface, serial input and output interface, voltage monitoring and other modules. Its structure is shown in Figure 1.

Figure 1. Structure diagram of the SPCE061A controller

2 Overall design of the smart car

The smart car is mainly composed of voice input circuit, voice output circuit, SPCE061A controller, drive circuit, etc. The SPCE061 controller of the car realizes voice control through the built-in microphone amplifier and the A/D channel and D/A channel with automatic gain function. The voice triggers the action of the car. After the car moves, the movement state of the car can be changed at any time through voice commands. If there is an obstacle in front during the movement, the car will stop automatically.

3 Hardware Design of Smart Car

1) Voice input circuit.

The voice input circuit of the car is shown in Figure 2. Among them, VM IC provides power for the microphone, VSS is the analog ground of the system, VCM is the reference voltage, and pins 1 and 2 are the positive and negative input pins of microphone X1 respectively. When speaking to the microphone, pins 1 and 2 will produce waveforms that change with the sound input from the microphone, and form two inverted waveforms at the two ports of SPCE061A, which are sent to the operational amplifier inside the SPCE061A controller for audio amplification. The amplified audio signal is converted into digital quantity through the ADC converter and saved in the corresponding register, and then these digital audio signals are compressed, identified, played, etc.

Figure 2 Voice input circuit

2) Voice output circuit.

The voice input circuit of the car is shown in Figure 3. Among them, VDDH is the reference voltage and VSS is the analog ground of the system. The audio signal is sent to the 9th terminal of the circuit by the DAC pin output of SPCE061A, and is sent to the integrated audio power amplifier SPY0030 through the adjustment end of the volume potentiometer R9. After audio amplification, the audio signal is output from SPY0030 and plays the sound through the external speaker connected to the J2 port.

Figure 3 Voice output circuit

3) Photoelectric detection circuit.

The photoelectric detection circuit of the car uses the E18-D80NK photoelectric sensor, which integrates transmission and reception. The infrared transmitting tube transmits infrared rays in a certain direction. When encountering an obstacle, the infrared rays are reflected and received by the receiving tube, so as to judge whether there is an obstacle in front of the car. The sensing distance of the obstacle can be adjusted according to the requirements through the fine-tuning knob on the sensor. A lens is added to the front end of the sensor to detect objects at a long distance by focusing. The sensor integrates amplification, comparison, and modulation circuits, so that the sensor is less affected by visible light. The connection diagram of the photoelectric detection circuit is shown in Figure 4.

Figure 4 Photoelectric detection circuit

4) Driving circuit.

The driving circuit of the car is a full-bridge driving circuit (Figure 5). The four transistors Q1, Q2, Q3, and Q4 form four bridge arms. Q5 controls the on and off of Q2 and Q3, and Q6 controls the on and off of Q1 and Q4. The driving circuits are used for the rear wheel power driving circuit and the front wheel direction driving circuit respectively. When pin 1 is high and pin 2 is low, Q1 and Q4 are turned on, Q2 and Q3 are turned off, and the motor drives the wheels to rotate; when pin 1 is low and pin 2 is high, Q1 and Q4 are turned off, Q2 and Q3 are turned on, and the motor drives the wheels to rotate in the opposite direction.

Figure 5 Driving circuit.

Connect ports 1 and 2 of the voice input circuit to the MICP and NICN pins of the SPCE061A controller respectively; connect port 9 of the voice output circuit to the DAC1 pin of the SPCE061A; connect ports 1 and 2 of the rear wheel power drive circuit to the IOB8 and IOB9 pins of the SPCE061A, and connect ports 1 and 2 of the front wheel direction drive circuit to the IOB10 and IOB11 pins of the SPCE061A; connect the OUT end of the photoelectric detection circuit to the IOB12 pin of the SPCE061A. The overall connection of the smart car is shown in Figure 6.

Figure 6 Overall connection diagram of the smart car.

4. Software Design of Smart Car

The software system of the smart car mainly uses voice recognition technology to control the automatic driving of the car, thus realizing the interaction between people and cars without manual operation. The intelligent control algorithm is shown in Figure 7. The software is designed in C language, which makes the program code concise and easy to read. In addition, the program design also uses the timing interrupt technology of SCPE061A. When the photoelectric detection circuit detects an obstacle ahead or a stop command is triggered, SCPE061A immediately performs an interrupt process to stop the car.

Figure 7 Block diagram of the intelligent algorithm.

The core of the software design of the smart car lies in speech recognition, which is mainly divided into training and recognition (Figure 8). In the training stage, the SCPE061A controller first performs analog/digital conversion, pre-emphasis, automatic gain control and other processing on the speaker's voice to realize the digitization of the voice digital signal, and then extracts the features of the processed voice signal to establish a voice feature model. The training process is mainly completed by the BSR_Train() function in the voice function library bsrv222SDL.lib. In the recognition stage, the SCPE061A controller performs the same analysis and processing on the collected speech, extracts the feature information of the speech, and then compares this feature information with the existing feature model. If the two reach a certain degree of match, the input speech is recognized. The recognition process is mainly completed by the BSR_InitRecognizer(), BSR_EnableCPUIncato r(), BSR_GetResult(), and BSR_StartRecognizer() functions in the speech function library bsrv222SDL.lib.

Figure 8 Speech recognition program flow chart

5 Conclusion

In the experiment, the correct recognition rate of the smart car was over 90%. During the experiment, it was found that the factors that affect the normal recognition of the car mainly include the noise of the surrounding environment and the distance between people and the car, which need to be improved in the future. This voice-controlled smart car robot can not only serve people in the future, but also replace people in a variety of occasions that are not suitable for human operations. Therefore, this voice-controlled car robot has important academic research value.