| As an important material,wood is widely used in all aspects of life.Cellulose,hemicellulose and lignin are the main components of wood.There are a lot of-OH in cellulose,and-OH is a hydrophilic group.Therefore,wood has affinity for water vapor and liquid water.Due to the loose and porous characteristics of wood,wood has certain hygroscopicity in the environment,which leads to the change of moisture content of wood and wood products with the change of temperature and humidity in the environment.When the moisture content of wood is greater than 30 %(fiber saturation point),the performance of wood is not affected by the change of moisture content.When the moisture content is less than 30 %,the dimensional stability and electrical conductivity of wood will be affected.Therefore,the moisture content management and control of wood products is particularly important.The drying method is a traditional method for detecting the moisture content of wood.Although the measurement results are more accurate,the measurement time is longer,and it is not a long-term measure in the application of the wood industry.In addition,the conductivity method,microwave method and ray method are also common methods,but the comprehensive cost,detection time and other factors are not suitable for large-scale application.In order to meet the needs of the development of modern wood industry,in recent years,some scholars have carried out various studies on the prediction of wood moisture content.Among them,near-infrared detection takes a short time,and the acquisition spectrum can be completed in just a few seconds.The high accuracy of measurement and the ability to achieve non-destructive testing are favored by scholars.The spectral coverage of the universal infrared spectrum is 780-2526 nm,but it is expensive.The five absorption bands of the moisture absorption spectrum are located near the wavelengths of 760,970,1145,1450 and 1926 nm,respectively,and the absorption peaks near 1450 and 1926 nm are more obvious.There are relatively few near-infrared spectroscopy acquisition devices for these two special wavelengths.To solve the above problems,this study aims to design a micro near infrared spectrum acquisition device with low cost,easy to carry and real-time detection of wood moisture content.Based on the characteristics of near infrared spectrum recognition technology,combined with partial least squares(Partial least squares method,PLS),a model is established to effectively achieve rapid detection of wood moisture content.Research contents and innovations:(1)According to the obvious spectral characteristics of water near 1450 and 1926 nm in the near-infrared region,two micro-spectral sensors C14272 and C14273 of Hamamatsu were selected as the core optical components in this experiment.(2)Considering the cost of the device and the design concept of integration,ST ’s stm32f107vct6 is selected as the main control chip.According to the requirements of the main control chip,the microcontroller module circuit is designed to realize the transmission and acquisition of spectral data.The AD5541 of ADI company is used as the digital-to-analog conversion chip,and the adjustable constant voltage source module circuit is designed to realize the control of the tunable filter inside the spectral sensor.The signal acquisition module is designed to collect the output signal of the sensor by using ADS8320 of Texas Instruments(TI)as the analog-to-digital conversion chip.Finally,according to the design scheme of each module,the schematic diagram and PCB diagram of the device are drawn.(3)According to the principle of separation of driver and application,the device driver and application were designed based on Free RTOS real-time operating system,including adjustable constant voltage source driver design,temperature acquisition module driver design,digital-to-analog conversion module driver design and communication module driver design.According to the demand of wood spectral acquisition,a multi-task system was established to determine the task execution logic and realize the rapid and accurate acquisition of spectral information.The experimental platform is built to test the function of the hardware module of the device.The test results are as follows.The output voltage accuracy of the adjustable constant voltage source circuit is higher than 0.76 %,the peak value is 48 m V,the accuracy of the temperature acquisition circuit is higher than 0.12 %,and the accuracy of the digital-analog acquisition circuit is higher than 0.82 %,which basically meets the design requirements of the peripheral circuits of C14272 and C14273 spectral sensors.The moisture content of oak veneer was selected for testing.The near-infrared spectra of wood samples with different moisture contents in the 1350 nm ~ 2150 nm band were collected,and the collected spectral data were modeled and analyzed by partial least squares method.Compared with the measurement data by the traditional drying method,the results show that the measurement value is good,the maximum error is 0.9 %,the minimum error is 0.3 %,and the overall error is controlled within 1 %. |