Abstract:    As the performance of technological innovation subsidy has been of great immediate significance, it has gradually become an important realistic problem and issue. Under the background that innovation is going at the head of the development of global economy, it has attracted the attention of the whole world. The paper analyzed the deep reasons and the relative mechanisms which are the determining factors of the performance of technological innovation subsidy by using the non-cooperative game theory under different conditions and situations. Then it has been shown that the repeated cooperative game between the managers of a number of the enterprises and a few of the relative government officials has attained the balance mainly because of information asymmetry, self-interest tendency of enterprises and lack of supervision mechanism to the associate parties. The main reason is that the above officials who have the powers and duties of approval, examination and supervision could not be competent when they complete the normal mission of the State and the government. Just on account of that, the performance of technology innovation subsidy is not good or even cheating constantly in the meanwhile. And then they could even give rise to the insufficiency of innovation ability of the micro-enterprises, medium-sized industries and further macro economy of the whole country in the cause of time. 
    As a result, this paper put forward two main innovative ways and ideas to smooth away the above-mentioned difficulties and all kinds of troubles, even straitened circumstances. The first one is that at the static level, we hold the view that the best regulatory mechanisms to eradicate persistent diseases are that the scientifically and validly incentive mechanism should be adopted as soon as possible, only if most of the subsidized enterprises are part of the low-efficiency category of technological innovation activities. And in the meantime, the audit strategy of regulatory authority ought to be implemented promptly if both audit cost and optimal reward amount are in the lower level at this time. And yet, at the dynamic level, the relative policy-making department should try to set up the three-party non-cooperative repeat game mechanism, the dynamic evaluation system of subsidy performance, and the supervision and governance mechanism of multi-party coordination on and on in order to prevent the government from suffering any other additional losses. As a result, most of these losses are due to enterprises′ cheating for subsidies and any other inefficient, waste, inaction, and misbehavior activities of all of the above mentioned. 
   The main innovation of this paper is that the Dynamic Cooperative Game Theory is used to analyze the internal mechanism of illegal activities, such as cheating for subsidies, which affect the performance of technological innovation subsidy. This paper constructs some models to ensure the optimal performance of technological innovation subsidy and the maximization of enterprise and government objectives, makes an in-depth analysis of the optimal static and dynamic strategies of the two independent audits, and puts forward corresponding concrete suggestions.

Key words: enterprise technology innovation, subsidy performance, cooperation game, cheat for subsidy, regulatory governance