In order to implement the declaration of the "Catalogue for Guidance of China’s High-Tech Products (2009)" and to meet the Shenyang government’s requirements on accomplishing the ’Eleventh Five-Year’ development plan which focusing on pollution reduction and environmental issues need to be resolved, the problem of treatment and soft measurement of the tar wastewater in the straw gasification combustion process is discussed in this work.Straw gasification technology is to put agricultural and forestry wastes under hypoxic or anaerobic conditions by thermal chemical reaction to generate combustible gases of CH4, CO, H2. Ever since the construction of the straw gasification stations in the city’s agriculture-related areas, the problems of water and air pollution caused by the cluttering and burning of the straw are improved much, and use efficiency of biomass energy is enhanced also. But with the repeated utilization of the straw gasification station, there are two main problems left to be solved. First, there is a need to offline measure the tar and chemical oxygen demand (COD) et al. during the straw gasification process, makes it difficult to timely manner detect abnormal pollutants and optimize the process parameter. Secondly, the problem on how to dispose the circulating spray liquid has become a bottleneck in the development and application of this technology. This paper focuses on the above two aspects.The work of this paper includes the following three aspects:(1) Design of wastewater treatment experiment. For stable operation of straw gasification equipment, there is a need to relax the requirements for water quality before filtration. Further, due to cost considerations, there is a need to determine the filter, the filtering time and filter types according to the experiment results. Based on these results, a wastewater treatment plant is designed, while the demand for input water quality of the wastewater treatment experiment is deduced.(2) Analysis of wastewater quality and establishment of wastewater soft measurement model. According to the requirements for the input water quality, there is a need to monitor and control the straw gasification tar, COD content in the straw gasification process. Thus soft measurement of water quality becomes a non-trivial problem. Because of the effects of Closed-loop control, material balance and other reasons, each gasifier temperature within the straw gasification furnace as well as the water quality indicators are highly-corrected (i.e., collinearity), there will induce large estimation error by applying traditional regression model, which makes it difficult to build the water quality soft sensor model. It is vital to study the effective wastewater quality soft sensor model with the input and output collinearities and the existence of non-linear relationship taken into consideration. In this work, the water quality soft sensor models using the OLS (Ordinary Least Square), PCR (Principal Component Regression), PLS (Partial Least Square), NNPLS (Neural Networks Partial Least Square) methods are established and the results using them are compared and analyzed. From the application results, we can maintain that the NNPLS method can obtain more accurate estimates of the water quality.(3) Process parameter setting method based on the similarity of PLS models. Taking into account the difference of raw materials in type, size and humidity, it is necessary to give optimized process settings for the newly added raw materials. To solve this problem, the PLS model which describes the temperature and water quality indicators are initially established first, and the similarities of these PLS models are utilized to give the set-points of the furnace temperature, air flow and other process indicators. |