Solid fuel heat value measurement method

Abstract

The invention discloses a solid fuel heat value measurement method which comprises the specific steps: (1) choosing varieties of calibration solid fuel samples with different heat values and respectively obtaining plasma emission spectrums of the solid fuel samples by utilizing a dot matrix measurement mode of an LIBS analysis system; (2) performing normalization processing on spectroscopic data matrixes and then utilizing K-CV to perform support vector machine related parameter optimization based on heat values on spectroscopic data; (3) utilizing SVM and optimal parameters to construct a heat value calibration regression model for normalization processed spectroscopic data matrixes; (4) utilizing the LIBS analysis system to obtain spectroscopic data matrixes of samples to the measured and inputting the normalization processed spectroscopic data matrixes to be measured into the calibration regression model to obtain heat values of the sample to be measured. The solid fuel heat value measurement method disclosed by the invention simplifies an analysis process in solid fuel heat value measurement and improves accuracy and precision of the LIBS applied to solid fuel heat value measurement.

Description

technical field [0001] The invention relates to the field of industrial technology, in particular to a method for measuring the calorific value of solid fuel. Background technique [0002] The calorific value of solid fuel is an important indicator for evaluating fuel quality. For example, calculation of fuel consumption, heat balance of combustion equipment, thermal efficiency, and determination of the mixing ratio of different fuels will all use the calorific value parameters of solid fuel. At present, the calorific value of solid fuel is mainly calculated by measuring the enthalpy difference between the reactant and the product with a cartridge calorimeter. This method is simple to operate and the result is more accurate, but because of the need for a cartridge calorimeter, the entire analysis cycle generally takes several hours, and it is difficult to meet the requirements of rapid detection of the calorific value of solid fuels in industrial processes. Therefore, it is...

AI Technical Summary

Problems solved by technology

[0004] Artificial intelligence technology has been more and more applied in the prediction of solid fuel calorific value, such as artificial neural network method (ANN), SVM (support vector machine method), but most of them are based on industrial analysis data or elemental analysis data However, there is a problem that the rapid detection of the calorific value of solid fuel cannot be achieved; and the parameter optimization problem of the support vector machine model is rarely solved by using the SVM method, so there is a problem that the model cannot be generalized

Images