Smart sensors emerged in the digital era

At present, the international sensor field has formed a basic consensus on the definition of “SmartSensor”, but the Chinese translation method has not yet reached a conclusion. The term “smart sensor” used in this paper is a personal reference. The smart sensor has one or more sensitive functions in terms of its functions, and can complete signal detection, conversion processing, logic judgment, function calculation, two-way communication, internal self-checking, self-calibration, self-compensation, self-diagnosis, and Devices with some or all of the above features.

From the perspective of use, the accuracy, stability, and reliability of the sensor are crucial. For a long time, most of the research work has been focused on hardware, although people are constantly using new materials to develop sensitive devices, improving the manufacturing process of sensor chips to improve the quality of the chip and improving the linearity, stability, and output drift of the sensor through external circuit compensation methods. However, there is no fundamental breakthrough.

In the 1970s, the world-renowned achievements of microprocessors brought about a revolution in digitalization and played a huge role in the development of instrumentation. For example, the rapid development of the virtual instrument (VXI) in the 1990s made the microcomputer-based measurement and control systems all require sensors to provide data for making real-time decisions. With the improvement of the degree of automation of the system and the increase of complexity, the sensor's overall accuracy, stability, reliability and response requirements are getting higher and higher. Traditional sensors, because of their single-function, unacceptable performance, and failure to meet a variety of test requirements, use microprocessor-based smart technology for sensors. In the late 1980s, people applied micro-machining technology to sensors, resulting in a new concept sensor "SmartSensor" - smart sensors.

Achievable functions The function of the smart sensor is proposed by simulating the coordinated actions of the human senses and the brain, combined with the research and practical experience of long-term testing techniques. Is a relatively independent smart unit, its emergence of the original hardware performance requirements have been reduced, while the help of software can greatly improve the performance of the sensor.

1. Compound Sensitive Function ---- We observe the natural phenomena around us. Common signals include sound, light, electricity, heat, force, and chemistry. Sensing element measurements generally take two forms: direct and indirect measurements. The smart sensor has a complex function that can simultaneously measure a variety of physical and chemical quantities, giving information that can more fully reflect the law of motion of matter. For example, the compound liquid sensor developed by the University of California, USA can measure the temperature, flow rate, pressure and density of the medium at the same time. A composite mechanical sensor developed by EG&GICSensors of the United States can simultaneously measure three-dimensional vibration acceleration, velocity, displacement, etc. at one point of an object.

2. Self-compensation and calculation functions ----The engineers and technicians engaged in the development of sensors for many years have been doing a lot of compensation work for the sensor's temperature drift and output nonlinearity, but they have not fundamentally solved the problem. The self-compensation and calculation functions of smart sensors open up a new path for sensor temperature drift and nonlinear compensation. In this way, to relax the sensor processing precision, as long as the repeatability of the sensor can be ensured, the signal of the test signal is calculated by the microprocessor through software, and the drift and the nonlinearity are compensated by the multiple fitting and the difference calculation method. Get more accurate measurements.

3. Self-checking, self-calibration, self-diagnosis function ---- ordinary sensors need regular inspection and calibration to ensure that it is sufficiently accurate in normal use. These tasks generally require that the sensor be disassembled from the site of use to the laboratory or inspection department. get on. If the online measurement sensor is abnormal, it cannot be diagnosed in time. The situation with smart sensors is greatly improved. First, the self-diagnosis function performs self-test when the power is turned on, and diagnostic tests are performed to determine whether the components are faulty. Secondly, it can be calibrated on-line according to the time of use. The microprocessor uses the metering characteristic data stored in the EPROM for comparison and proofreading.

4. Information storage and transmission ---- With the rapid development of the smart distributed control system (SmartDistributedSystem), the smart unit is required to have the communication function and use the communication network to conduct two-way communication in digital form, which is also one of the key signs of the smart sensor. Smart sensors implement various functions by testing data transmission or receiving instructions. Such as gain setting, compensation parameter setting, internal inspection parameter setting, test data output, etc.

Integration of smart sensors The development of large-scale integrated circuits allows sensors and corresponding circuits to be integrated on the same chip. This kind of sensor with some intelligent functions is called integrated smart sensor. The function of integrating smart sensors has three advantages. : Higher signal-to-noise ratio: The weak signal of the sensor is first amplified by the integrated circuit signal and then transmitted over a long distance, which can greatly improve the signal-to-noise ratio. Improve performance: Because the sensor and the circuit are integrated on the same chip, zero drift, temperature drift and zero position of the sensor can be automatically calibrated by the self-calibration unit on a regular basis, and an appropriate feedback method can be used to improve the frequency response of the sensor. Signal conditioning: The analog signal of the sensor is normalized by a programmable amplifier, and converted to a digital signal by analog to digital. The microprocessor performs digital specification in several forms of digital transmission, such as serial, parallel, frequency, Phase and pulse.

The manufacturing base of micro-machining technology and software smart sensors is based on micro-machining technology. Silicon is mechanically, chemically, and weldingly processed, and then encapsulated with different packaging technologies. In recent years, a LIGA process has been developed (deep X-ray). Lithography electroplating into a sensitive film) is used to make sensors.

Smart sensors generally have strong real-time capabilities. In particular, dynamic measurements often require data acquisition, calculation, processing, and output in microseconds. A series of functions of the smart sensor are performed with the support of the program. Such as the number of functions, basic performance, ease of use, and reliable work, most of them rely on software design and its quality to some extent. There are mainly five categories of these software. Including scale conversion, digital zero adjustment, nonlinear compensation, temperature compensation, digital filtering technology. The state of application at home and abroad more than ten years ago, the United States Honeywell company developed the first smart sensor. It combines the silicon sensing element technology with the microprocessor's calculation and control capabilities to establish a new sensor concept. At present, smart sensors are mostly used for the measurement of pressure, force, vibration impact acceleration, flow rate, and temperature and humidity. Such as the United States Honeywell's ST3000 series of smart transmitters, Germany Strohrmann's two-dimensional acceleration sensors.

At present, smart sensor systems are digital, but their communication regulations still use 4 to 20 mA standard analog signals. International standardization research institutes are actively launching international standards for digital standards (field bus). In the current transition phase, the HART protocol (Highway Addressable Remote Transducer) is used. This is a communication protocol for smart sensors that is compatible with existing 4-20 mA systems. Both analog and digital can communicate simultaneously. This makes the products of different manufacturers universal.

The research of smart sensors in China mainly focuses on professional research institutes and universities. It began in the mid-1980s. In the late 1980s, China National Defense Science and Technology University, Beijing University of Aeronautics and Astronautics, Zhejiang University and other universities and colleges successively reported research results. In the early 1990s, several domestic research institutes successfully developed practical smart sensors using hybrid integration technology, marking the research of smart sensors in China has entered the international ranks, but compared with foreign advanced technologies, we still have a big gap . Mainly in: 1, advanced calculation, simulation and design methods, 2, advanced micro-machining technology and equipment, 3, advanced packaging technology in equipment, 4, reliability technology research and other aspects. Therefore, strengthening the research of technology and introducing advanced equipment to improve the overall level is our future direction.

In the coming years, smart sensors will be expanded into fields such as chemistry, electromagnetics, optics and nuclear physics. It can be foreseen that more and more smart sensors will play a role in various fields of our national economy.