Through experiments, we designed an intelligent noise reduction filter window that can effectively isolate most of the noise. The noise reduction window is composed of a noise reduction device, a double-layer glass window, a lifting and folding noise reduction panel and a solar panel. The product controls various functions of the noise reduction device through a smart home APP. APP takes the mobile terminal of the mobile phone as the center, establishes the connection between the mobile terminal of the mobile phone and the device through WiFi network, and realizes the functions of specific noise reduction of external noise, lifting of noise reduction folding board, viewing of solar panel power and real-time uploading of decibel size of external environmental noise. The noise reduction device sends the monitored ambient noise data to the mobile phone APP end, and then uploads it to the background big data statistics and management platform. Through the effective analysis of the data, the platform can help government departments to supervise and manage the areas with serious noise pollution. The innovation of the product lies in the combination of solar cell power supply and button cell power supply, which makes the life of the noise reduction device imitating the human ear structure increase, avoids the defects of the two power supply methods, and achieves the complementary effect of advantages. At the same time, the use of solar photovoltaic glass for direct conversion of light and electricity, more energy saving and environmental protection. The key to product technology is active noise reduction technology and prediction algorithm. Eliminate ambient noise through active noise reduction technology. The noise reduction effect of the product is far better than that of ordinary noise reduction equipment on the market, which can better create a quiet indoor environment for users. Adding the prediction mechanism over time can find the optimal weight of the filter faster in the case of insufficient data in the early stage, speed up the convergence and tracking ability of the algorithm, and improve the response speed of the Fir filter.