Abstract: The launch array of the active towed array, in the form of horizontal collinear, experiences interference from seabed reflection waves during operation. This seriously affects the equipment's echo detection performance. To address this issue, we improved the delay estimation accuracy of echo detection to enhance equipment performance under submarine reflection interference. Using Richardson-Lucy (R-L) deconvolution algorithm, we first analyzed and deduced the convolution model of matched filter and then provided an iterative solution formula for underwater acoustic channel impulse response function. We analyzed target echo detection performance under different iteration times, input signal-to-noise ratios, and seabed reflection wave interferences while conducting relevant simulation tests. Simulation results indicate that R-L deconvolution algorithm has a narrower main lobe width and lower sidelobe level making it suitable for low SNR conditions with faster convergence rates. The active towed array sonar in horizontal collinear form shows significant improvement in its detection performance while verifying once again the applicability of our algorithm through experimental data processing on a lake.