Sensor "easily encounter" intelligent manufacturing

With the promotion of Industry 4.0, smart manufacturing and the Internet of Things, sensors will be further applied and perfected, and they will also face significant problems.

A common difficulty in the field of industrial sensors is that after the sensor is used for a period of time, the sensor is affected by factors such as product aging and parameter drift. The detection will inevitably result in errors and it is difficult to adjust the correction. For this reason, many product managers and R&D engineers pay innumerable sweats, but the effect is obviously unsatisfactory.

On the other hand, with the promotion of Industry 4.0, smart manufacturing and the Internet of Things, sensor applications will be ubiquitous, and the problem of errors will become increasingly prominent. At the same time, new application areas also put new demands on sensors, and must be further improved in product integration, wired and wireless interface protocols, and power consumption performance, in order to increase sensor flexibility and cost advantages.

Seeing the leopard in the tube, a glimpse of the new trend of industrial sensors

Industry 4.0 is a wide-ranging concept. It mainly uses information systems to digitize the supply and manufacturing processes in the production process to facilitate the management, analysis, and promotion of the production process so as to increase the level of intelligence in the manufacturing industry. Therefore, in the production process, the importance of the sensor neural network is naturally self-evident.

The detection accuracy and performance of the sensor network directly affect the efficiency of production. Wang Sheng Yang, general manager of Nash Microelectronics, said that more and more integrated solutions began to be used in large numbers, mainly in the following two aspects. Firstly, use ASIC instead of discrete circuit to make the sensor have better portability and reduce cost. For example, pressure transmitters before TI all use discrete circuits, BUF, ADC, and processor are all independent devices; TI The newly introduced solution is a complete ASIC chip that integrates all logic functions. Second, multiple sensor ASICs begin to integrate horizontally. For example, industrial field temperature control solutions originally use independent circuits and outputs such as RDD and thermocouples. 32 channels can be measured at the same time. Each channel can be combined to achieve various measurements such as RDD, thermocouple, and diode. It is very convenient to manage.

On the other hand, there is a strong correlation between productivity and the ability to work together in sensor devices. However, at this stage, the sensor is still mainly based on analog signal output. It can only realize one-way flow of information, and cannot adjust and configure the sensor. The amount of information that can be carried by the analog output is very limited, and only sensor detection information can be sent. The sensor's own diagnostics, ID, and operating status are difficult to transmit. Therefore, the digital interface is an inevitable trend of subsequent development, such as: HART protocol, CAN bus, point-to-point IO_link bus, etc., can achieve two-way communication, real-time configuration of the sensor state.

"Power consumption is a problem that the sensor network must face and improve." Wang Sheng Yang continued, because the sensor network system will become larger and larger, and the power consumption of each sensor device will add up, which means huge cost expenses. In addition, many industrial applications are still battery-powered, and battery life is difficult to guarantee. Therefore, in sensor devices, sensors, signal processors, power management, MCUs, and wireless communications all need to implement low-power processing to improve the equipment's endurance.

Therefore, the glimpse of the leopard in the tube is evident. Integration, interface, and power consumption are just the tip of the iceberg of sensor development. In the future, more sensors will work closely with algorithms in vertical fields to jointly promote the development of intelligence.

Each display supernatural, sensor error solution

In industrial production, small errors in sensor detection are likely to result in huge losses for enterprises. This is a situation that companies are very reluctant to see. So, how can the error of the sensor be avoided? The following is a case study of sensor industry experts in product applications, and may provide a lot of reference for everyone.

Smart meter reading brings a very convenient experience, but there are also many problems. Dong Shenglong, deputy general manager of Xinao, said that if the data integrity and timeliness of meter reading are problematic, the experience will be very poor. This is often caused by multiple reasons. The first is that the atmosphere sensor can only be identified and cannot be analyzed quantitatively well; the second is that the structural design and production process are not mature, resulting in defects in consistency; the third is the lack of active sensors for the meter. Most of the electricity meters from the factory to the application require on-site gas charge correction compensation coefficients. Xinao’s solution mainly takes the following steps:

The first is to study the measurement principle of component sensors and form a dynamic compensation scheme for metering. Through a large number of experiments, the problem of component adaptability is solved.

Second, through a large number of experimental verifications, a real-time tracking and on-line correction scheme is established between natural gas characteristics and sensor metering output. The recognition time is less than one minute, and 80% to 97% of methane content is automatically identified.

Third, on the premise of guaranteeing the speed, the design structure is optimized, the influence of the flow velocity on the sensor is reduced, and the measurement stability is improved;

Fourth, increase the self-detection capability of the instrument and perform real-time detection and alarm on the working status of the instrument.

Due to the ease of installation and the need for power and wiring, passive wireless sensors are well-received and can be used in many extreme environments. Li Zechen, general manager of Saihe Information Technology Co., Ltd., said: “In automotive exhaust gas treatment systems, the temperature range of exhaust gas is between 600 and 1200 degrees, which results in the scarcity of gas sensors capable of detecting exhaust gas such as overoxidation*, carbon monoxide, etc., and the interference of sensors. And the aging situation will be more serious.The surface acoustic wave sensor is one of the wireless passive sensors. In the process of aging and environmental interference changes, the method of setting the synchronization drift of two frequency points is used to detect the frequency difference between two frequency points. Physical quantity, so that accurate detection values ​​can be obtained.”

However, in different applications, the reasons for the error are different. Enterprises should categorize the source according to the actual situation and solve the problem of error.