This paper describes a wireless data transmission technique using acoustic waves as the information carrier for online injection molding process measurements, including melt temperature, pressure, velocity and viscosity. A design scheme of the optimization of the input electrical impedance matching with the transmission system of the multivariate sensor based on the bond graph modeling approach is proposed. Bond graph models of each constituent component within the transmission system are established, where parameters regarding the properties of the piezo-material, layer thickness, and input electrical impedance could be manipulated easily. The effects of different combinations of capacitor and inductor that constitute the input electrical circuit are investigated. An optimal combination for electrical impedance matching is determined through bond graph simulation. This optimal solution improves the design of sensor circuits with the capability of generating the highest output gain or increasing the transmission distance under the same electrical excitation, meanwhile ensuring that the received signal has high signal-tonoise-ratio (SNR) for further parameter retraction. Bond graphing has been shown as an effective approach for guiding the design of complex, cross-domain, and embedded sensing systems.