Natural gas hydrates are regarded as a vital strategic energy resource for the future owing to their high energy density and clean combustion characteristics. To facilitate research into the physical and mechanical properties of pressure-maintained hydrate samples, this paper presents an integrated multi-parameter online analysis system capable of rapidly measuring the P-wave velocity, electrical resistivity, thermal conductivity, and shear strength of core samples under pressure-maintaining conditions. The system comprises hardware acquisition boards based on ZYNQ and ARM platforms, specialized measurement probes, and comprehensive data acquisition and analysis software. To mitigate the susceptibility of P-wave signals to noise interference, an improved denoising algorithm combining Empirical Mode Decomposition (EMD) and wavelet thresholding is proposed. By employing autocorrelation function analysis, the algorithm identifies the transition boundary between noise-dominated and signal-dominated Intrinsic Mode Functions (IMFs), subsequently applying wavelet soft-thresholding to the noise-dominant components. Experimental results demonstrate that the proposed algorithm achieves a superior signal-to-noise ratio (SNR) compared to traditional EMD methods, particularly under low SNR conditions. System validation indicates measurement accuracies of 3.2% for P-wave velocity at 20 °C, 1.76% for electrical resistivity at 25 °C, and within 7% for both thermal conductivity and shear strength. Furthermore, sea trials conducted aboard the “HAIYANG SHIYOU 708” drilling vessel confirm that the system operates stably and effectively fulfills the requirements for deep-sea core parameter characterization.