Types and structures/characteristics/functions/implementations of smart sensors

The concept of smart sensors was first introduced by NASA in the process of developing spacecraft and was formed into a product in 1979. Spacecraft require a large number of sensors to send data such as temperature, position, speed and attitude to the ground or spacecraft. Even with large computers, it is difficult to process such a large amount of data at the same time. In addition, spacecraft limit the size and weight of computers. It is hoped that the sensor itself has information processing capabilities, so when the sensor is combined with a microprocessor, a smart sensor will appear.

Smart sensors are a type of sensor that can sense and detect information about an object, learn, judge and process signals, and have communication and management functions. Smart sensors have the ability to automatically calibrate, calibrate, compensate and collect data. Their capabilities determine that smart sensors have high accuracy and resolution, high stability and reliability, and good adaptability. Compared with traditional sensors, they have a high performance-price ratio.

Types and structures of smart sensors

1) The types of smart sensors are as follows:

A. The sensor has the ability to judge

B. Sensors with learning capabilities

C. Sensors with creativity

2) Structure of smart sensor

The intelligent sensor system is mainly composed of sensors, microprocessors and related circuits, as shown in the figure. The sensor converts the measured physical and chemical quantities into corresponding electrical signals, sends them to the signal modulation circuit, and then sends them to the microprocessor after filtering, amplification and A/D conversion.

The microprocessor calculates, stores, analyzes and processes the received signals, adjusts the sensor's feedback loop and signal conditioning circuit to achieve regulation and control of the measurement process; on the other hand, the processing results are transmitted to the output interface, and the interface circuit processes according to the output format, and the output interface customizes the digital measurement results. The microprocessor is the core of the smart sensor. Due to the functions of various software, the sensor is intelligent and the performance of the sensor is greatly improved.

Features of smart sensors

(1) High precision

Smart sensors eliminate the impact of unexpected errors and ensure the accuracy of smart sensors through automatic zeroing and division, real-time automatic comparison of standard references for overall system calibration, nonlinear system errors, and large amounts of real-time data analysis and processing.

(2) High reliability and high stability

The intelligent sensor system can automatically compensate for the operating characteristics and environmental parameter changes and ambient temperature sensitivity caused by zero drift, and the system power supply voltage fluctuation caused by drift; it can automatically change the measurement parameter range, perform real-time system self-check, analyze and judge the rationality of data for emergency processing, and automatically handle abnormal situations.

(3) High signal-to-noise ratio and high resolution

Smart sensors have data storage, memory and information processing functions. Through digital filtering and correlation analysis, they can remove noise from input data and automatically extract useful data. Through data fusion, neural network technology can eliminate the influence of many parameters under cross-sensitive conditions.

(4) Strong adaptability

The intelligent sensor has the functions of judgment, analysis and processing. It can make decisions on the power supply and data transmission rate of each part according to the working conditions of the system, using a high/top computer, so that the system can work in the best low-power state and optimize the transmission efficiency.

(5) Higher performance and price ratio

Smart sensors have high performance. Unlike traditional sensor technology, which is achieved through the pursuit of perfection, the sensor itself is "obtained" in every link of sensor design and debugging, and elaborate craftsmanship. Instead, it is achieved through the combination of computers and microprocessors/microchips, using integrated circuit technology and cheap and powerful software, so it has a high performance price ratio.

The role of smart sensors

The function of smart sensors is simulated by simulating the coordinated actions of human sensory organs and the brain, combined with long-term testing technology research and practical experience. It is a relatively independent intelligent unit. Its emergence reduces the stringent requirements for the original hardware performance and greatly improves the performance of sensors through software assistance. Smart sensors can usually achieve the following functions:

(1) We observe the natural phenomena around us. Common signals are sound, light, electricity, heat, force, and chemistry. Sensitive element measurements are usually measured in two ways: direct measurement and indirect measurement.

Intelligent sensors have complex functions and can measure various physical and chemical quantities at the same time. The information they provide can more comprehensively reflect the laws of material movement. As a composite liquid sensor from the University of California, temperature, velocity, pressure and density can be measured simultaneously. The composite mechanical sensor developed by EG&GIC Sensor Company in the United States can simultaneously measure the three-dimensional vibration acceleration, velocity and displacement of a point.

(2) Adaptive function: Intelligent sensors can automatically adjust their own characteristics to adapt to such changes under a certain range of conditions.

Adaptive technology can extend the service life of a device or component because it can compensate for parameter drift caused by aging components. At the same time, it also expands its working field because it can automatically adapt to different environmental conditions. Adaptive technology improves the repeatability and accuracy of the sensor. Because the correction and compensation values ​​are no longer average values, but the true correction values ​​of the measurement point.

(3) Ordinary sensors with self-test, self-calibration and self-diagnosis functions need to be checked and calibrated regularly to ensure that they are accurate enough in normal use. These generally require that the sensor be removed from the site to the laboratory or inspection department, and there is no timely diagnosis of whether the online measurement sensor is abnormal.

When using smart sensors, the situation is greatly improved. First, a self-diagnostic function is performed on the power supply, and diagnostic tests are used to determine the failure of the component. Second, online correction can be made based on the use time, and the microprocessor uses the measurement characteristic data in the E2PROM for checking and proofreading.

(4) Information storage is often the key to success. Smart sensors can store a large amount of information that users can query at any time. This information can include historical information about the device.

For example, how many hours the sensor has worked, how many times the power supply has been changed, etc. It also includes all the data and diagrams of the sensor, including configuration selection instructions, etc. In addition, the serial number, production date, catalog and final results of the final factory test are included. The content can be unlimited, only limited by the storage capacity of the smart sensor itself. In addition to the four major functions of process data processing, self-diagnosis, configuration and information storage, smart sensors also provide digital communication and self-adaptation capabilities.

(5) Data processing is a very important task in dealing with data processing, and smart sensors themselves provide this function. Smart sensors can not only amplify signals, but also digitize signals and use software to implement signal adjustment. Generally, basic sensors cannot give linear signals, and process control takes linearity as an important goal. Smart sensors can linearize nonlinear signals through the lookup table mode.

Of course, this data sheet should be compiled individually for each sensor. Another example of process data processing for smart sensors is filtering the digital signal by means of a digital filter, thereby reducing interference from noise or other relevant influences. It is much easier to develop complex filters using software than using discrete electronic circuits. Compensation for environmental factors is also an important task for data processing. Microcontrollers help to improve the accuracy of signal detection.

For example, by measuring the temperature of the basic detection element, the correct temperature compensation coefficient can be obtained to achieve temperature compensation of the signal. Nonlinear compensation and other more complex compensations can also be achieved using software. This is because the lookup table can generate curves of almost any shape. Sometimes several different physical quantities must be measured and processed in order to give their respective data. The sign controller of the smart sensor can easily realize the addition, subtraction, multiplication and division operations of multiple signals. In the process of process data processing, smart sensors can be very important.

In addition, it is also beneficial to move these operations from below the central control room to a point close to the point where the signal is generated. One is that the cost of sending additional signals to the control room is high, and smart sensors save the cost of additional sensors and leads. Another is to detect additional information at the point of application of the information, which greatly reduces the negative effects introduced by long-distance transmission (such as noise and potential difference), making the signal more accurate. The third is to simplify the software in the main controller and increase the speed of the control loop.

(6) Another main feature of smart sensor configuration function is the configuration function.

How many times should the signal be amplified? Is the temperature sensor output in Celsius or Fahrenheit? For smart sensor users, you can freely choose the configuration you want. For example, detection range, programmable pass/interrupt delay, select group counter, open/normally closed, 8/12-bit resolution selection, etc. This is just a small part of the countless configurations of today's smart sensors. Flexible configuration functions greatly reduce the need for users to develop and replace the necessary types and quantities of different sensors. Using the configuration function of smart sensors, the same type of sensors can work in the best state and can do different jobs in different occasions.

(7) The function of digital communication is as described above. Since smart sensors can generate a large amount of information and data, the single input and output of ordinary sensors cannot provide the necessary input and output for device data. But you cannot use one lead for each information because it makes the system very complicated, and it requires a flexible serial communication system.

In the process industry, point-to-point connections and serial networks are commonly seen. The general trend is towards serial networks. Since the smart sensor itself has a single-chip microcomputer and is digital, it can naturally be configured with digital serial communication for external connections. Because the serial network's ability to resist environmental influences (such as electromagnetic interference) is much stronger than ordinary analog signals. By matching serial communication with the device, the transmission of information can be effectively managed so that data is output only when needed.

4. Implementation of Smart Sensors

4.1 Implementation of Smart Sensors

Currently, with the development of sensor technology, there are three ways to implement smart sensors.

a. Use computer synthesis, i.e. intelligent synthesis;

B. Use special functional materials, i.e. smart materials;

C. Use functional geometric structures, i.e. smart structures. Smart synthesis is the combination of sensor devices and microprocessors, and is currently the main method.

4.2 Intelligent Sensor Technology

According to the synthesis method of sensors and computers, current sensing technology uses the following three specific methods to realize smart sensors.

Non-integrated module mode. Non-integrated intelligent sensors are intelligent sensor systems composed of traditional basic sensors, signal conditioning circuits, microprocessors and digital bus interfaces. This type of non-integrated intelligent sensor was developed under the rapid development of fieldbus control systems. The original complete set of production equipment of the automation instrument manufacturer remains basically unchanged, and a digital microprocessor bus interface board is added to assemble it, and it is equipped with intelligent software for communication, control, self-correction, self-compensation, and self-diagnosis to realize the functions of intelligent sensors. This is the most economical and fastest way to build intelligent sensors.

Integrated implementation. The smart sensor system uses micromachining technology and large-scale integrated circuit technology to manufacture sensitive elements, signal conditioning circuits and microprocessor units with silicon as the basic material, and integrates them on the chip. Integrated implementation enables the miniaturization and structural integration of smart sensors, thereby improving accuracy and stability. After the sensor element array is composed of an array and corresponding image processing software, graphic imaging can be realized to form a multi-dimensional image sensor. At that time, the smart sensor has reached its most advanced form.

Comprehensive Implementation

Hybrid implementation. There are many challenges in realizing smart sensor systems on a chip. Depending on the needs and possibilities, we can integrate all the integrated links of the system, such as the sensing unit, signal conditioning circuit, microprocessor unit and digital bus interface, and integrate them in different combinations on two or three chips.