Abstract : This paper applies the multi-ultrasound sensor data fusion technology to electronic blood pressure measurement and analyzes in detail the principle of measuring blood pressure by using multiple ultrasonic sensors to form a sensor array. On the DSP platform, the adaptive weighted estimation method of the multi-sensor data fusion algorithm is used to fuse the information collected by multiple sensors and upload the fusion result to the host computer. Experiments show that this method can better overcome the measurement error caused by the subject's physical activity. Keywords : ultrasound; data fusion; adaptive; weighting factor 1. Introduction Blood pressure reflects the functional status of the human heart and blood vessels. Blood pressure detection is of great significance for disease prevention and treatment and early detection of diseases. Blood pressure measurement often uses indirect methods. This paper focuses on the ultrasound method, which uses the Doppler effect of ultrasound on blood flow and blood vessel wall movement to detect systolic and diastolic blood pressure. Because the Doppler frequency shift has a relatively stable correlation with blood pressure, the blood pressure value measured by using the Doppler principle is relatively accurate. However, in the previous single-sensor measurement system, since the variance of the sensor cannot be changed, the only way to reduce the error is to increase the number of measurements and calculate the average value, which greatly increases the computational load of the system [2]. This paper studies an adaptive weighted fusion algorithm for multi-sensor data. It requires no prior knowledge of the sensor measurements and fuses the data to produce a fusion value with the minimum variance. The estimated variance is less than that obtained using the mean square estimation method for a single sensor and the average of the multiple sensor values. 2. System Structure Three sets of ultrasound probes with different frequencies are placed at different positions on the cuff. First, the microcontroller emits pulse signals of three different frequencies, which drive the ultrasound sensors through the transmitting circuit. Simultaneously, the microcontroller drives an air pump to inflate the cuff via a driving circuit. When the cuff pressure is higher than the blood pressure in the artery, the blood vessel is blocked, and the vessel wall remains stationary. Therefore, the ultrasound echo detected by the microcontroller does not exhibit a Doppler shift. When the cuff pressure is just below or above the blood pressure, the blood flow is relatively greater than the blood vessel movement, and the ultrasound echo detected by the microcontroller will exhibit a Doppler shift. During a cardiac cycle, as the cuff pressure gradually increases, the opening and closing time of the blood vessels gradually decreases until the opening and closing coincide. The pressure on the pressure sensor at this point is the systolic pressure. Conversely, when the cuff pressure decreases, the time between opening and closing increases until the closure of one blood vessel coincides with the opening of the next blood vessel. The pressure on the pressure sensor at this point is the diastolic pressure. Therefore, with each heartbeat, the blood vessels open and close, and the systolic and diastolic pressures can be detected by means of the waveform output by frequency shift [5]. After fusing the blood pressure data of each group through DSP data, more accurate blood pressure data can be obtained. The principle block diagram of the system is shown in the figure. Blood Pressure Measurement System Based on Multi-Ultrasound Sensor Data Fusion Technology Full Text Download