Smart network car market competition will be very intense, the market does not need so many companies or start-up companies, some stay and grow, some are acquired, and some are slowly dying. Smart network car is a kind of smart car. Smart car is a relatively large concept, divided into autonomous and network-based, autonomous is only rely on vehicle sensors for perception of smart cars, network-type both rely Sensor on car sensors, but also rely on V2X communication to achieve interconnection, combined to become Intelligent network of cars. Today, the network is an inevitable trend, so the intelligent network of cars and smart cars have no distinction, the intelligent network car on behalf of the basic smart car. From a primary concept to the technical architecture and development of the route to build, until now, intelligent network of cars has been completed in the country a round of basic planning. Intelligent network of vehicles equipped with advanced automotive sensors, controllers, actuators and other devices, and integration of modern communications and network technology to achieve the car and X (people, cars, roads, clouds, etc.) intelligent information exchange, sharing, with complex environmental awareness , Intelligent decision-making, collaborative control and other functions, can achieve 'safe, efficient, comfortable, energy-saving' driving, and ultimately to achieve alternative people to operate a new generation of cars. '
Car network refers to the triple play. First, the car has a local area network, CAN / LIN bus, there may be Ethernet later. Second, between the car and car, between the car with the road, forming a car network. Third, the overall mobile Internet, 3G, 4G, 5G. The three together constitute Fuel Rail Pressure Sensor a car network. And for smart network car, you can understand it as a car network and intelligent car intersection. And for autopilot, it is the core part of the intelligent network car in the driving function, the ultimate stage is to achieve unmanned driving.
The development of intelligent network car is an automated or intelligent dimension, that is usually mentioned in the international level 1-5, driving support to unmanned driving. Information services the most basic level, including navigation, vehicle location data upload, OBD and other data upload. Synergy perception: The vehicle can communicate with the signal lights and communicate with the blind area vehicle. For example, Audi to the end of the introduction of the traffic lights auxiliary system, through the 4G network to traffic signal information to the Audi car, the vehicle can automatically adjust Speed Sensor the engine according to the signal phase stop, and provide red light warning function. Collaborative decision-making and control: part of the decision-making information from the cloud to complete, the cloud can achieve scheduling, decision-making, the vehicle will become an executive body. But at present in the country, the network has just started.
Environmental awareness is the foundation to help the vehicle understand environmental information, including traffic rules, which involves the sensor itself, such as image recognition, radar, information fusion, V2X, etc., because we can now think that V2X is a kind of environmental perception. At the sensor level, China's camera sensors have basically achieved autonomy, the volume is relatively large; ultrasound, long distance to do less, mainly rely on some mainstream suppliers; millimeter-wave radar is currently hot in the country, but the face The problem is that the price of this sensor has been very low, about a few hundred, domestic manufacturers Throttle Position Sensor to product stability, reliability and price with the manufacturers have fierce competition; laser radar, some domestic start-up companies to do, but the current possession The rate is still relatively low.
Decision-making involves decision points from point A to point B, how to avoid obstacles, choose paths, or stop waiting. At present, the decision-making level is the smallest gap with foreign countries, this part of the team using the depth of learning algorithms for end-to-end decision-making, some based on traditional rules for path planning, the overall gap is not large.
Control the implementation of the decision command into the actual speed of the vehicle, the steering wheel angle and so on. Which involves the design of the controller, the implementation of the underlying mechanism of the remote control system, and the implementation Pressure Switch of this piece has been our short board, so the future in the chassis control system is still a lot of room for development, through the intelligent network The core technology of the car, environmental perception, decision making, control the implementation of three blocks.