Scholars have become increasingly aware of the systematicness and complexity of the networks. To determine the characteristics of a cooperative innovation network, they have started using a system dynamics method, complex network theory, and social network analysis methods to further explore the mechanism of innovation networks [

39]. In 1999, Barabasi and Albert were the first to propose a classic BA scale-free network model [

40]. This made the generation and evolution of complex networks a popular topic. The most important contribution of the BA model is the proposal of two critical evolutionary mechanisms, the growth of the network, and the preferential connection. The fitness model [

41] and the local-world evolving network model [

42] have been subsequently introduced.

#### 4.3. Simulation Steps

As each stage of the enterprise requires different networks [

49], the government plays an essential role in the early stage of network development. However, during the advanced stage, government involvement gradually declines, while industry associations play a major role [

50]. Therefore, we considered the changes to innovation behavior in the initial and mature stages.

**Step 1.** It is assumed that the initial market is empty and the parameters are set, as follows. Assume the types of resources (including a series of innovative resources such as human resources, information, knowledge, and technology) ${k}_{0}=20$ in the initial stage of the innovation network.

As the innovation resources will not be static with the development of science and technology, the types of resources continue to increase, assuming that each period of resources increases by one probability, ${r}_{s}=0.5$.

It is assumed that the number of enterprises in the initial network ${n}_{0}=30$ and the total number of simulations $T=100$. It is assumed that the proportion of the initial period $r=0.3$, according to the theory of technological innovation diffusion curve.

As the newly entered enterprise will have a tendency to connect with the network subject, for n_{t} enterprises that enter the market, each one will have 1–3 resources. The initial ownership of each resource is subject to the mean value ${\mu}_{0}\cdot (1+\omega )\cdot t$, and the mean and variance of the initial possession of each resource are ${\mu}_{0}=1$ and ${\sigma}^{2}=0.2$, respectively.

The initial resource mean value increases with time, $\omega =0.2$.

Recently entered enterprises attempt to establish contact with those that are generally rich in resources, but these rich enterprises are not necessarily willing to cooperate with the new ones. Assume that the number of enterprises with which each resource-rich enterprise is willing to cooperate is c = 5 each time. For enterprises, there are two motivations for cooperation: the object has more resources than it does and the object has resources that it does not. The resources that correspond to these two motivations are dissimilar, and the latter should be more significant. Therefore, it is assumed that the discount of the first resource relative to the second resource is z = 2.

At maturity, government support is weaker than during the initial period. Assume that the initial period ${s}_{1}=0.3$, the mature period ${s}_{2}=0.1$, and resource learning efficiency ${\omega}_{s}=0.1$.

The types of government support objects are indicated by f (variable: Policy 1 indicates support for newly entered enterprises, policy 2 indicates support for enterprises with less resources, and policy 3 indicates random support).

**Step 2.** For a certain period $t$$(t=0,1,\dots ,T)$, determine the following. If $t=0$, the number of enterprises that are expected to enter the market, $({n}_{t})$ is ${n}_{0}$. Since universities in the early-stage period provide more scientific and technological resources, more enterprises are developing than in the maturity period.

Therefore, if $1\le t<T\cdot r$ (initial period), the number of enterprises entering the market $({n}_{t})$ is a random number between 6 and 10.

If $T\cdot r\le t\le T$ (mature period), the number of enterprises that are entering the market $({n}_{t})$ is a random number between 1 and 5.

The absorptive capacity of the newly entered enterprises b_{i} is directly proportional to the total amount of resources that are initially owned. To facilitate calculation, it is assumed that the absorptive capacity remains unchanged with time. This hypothesis does not affect the nature of evolution. Add n_{t} enterprises to the market.

**Step 3.** Establish cooperation using Equations (1)–(3). Subsequently, according to the type f of government support, choose the supporting enterprises. If $f=1$, support $({n}_{t})$ enterprises entering the market in the current period. If $f=2$, support enterprises with the lowest 20% of the total resources in the market at that time. If $f=3$, randomly select 20% of the enterprises whose total resources are not in the top 20% to support.

If enterprise i is supported by the government, then the probability ${P}_{i-{j}_{l}}$ of establishing a cooperative relationship between enterprise i and enterprise ${j}_{l}$ is defined as ${P}_{i-{j}_{l}}={P}_{{j}_{l}-i}=({P}_{i}^{{j}_{l}}+{P}_{{j}_{l}}^{i}+s)/2$ where s is the government’s support. If $1\le t<T\cdot r$ (initial period), $s={s}_{1}$. If $T\cdot r\le t\le T$ (mature period), $s={s}_{2}$. If ${P}_{i-{j}_{l}}>1$ , ${P}_{i-{j}_{l}}=1$.

Based on this probability, whether the two enterprises cooperate is determined. If cooperation occurs, then the cooperation time ${y}_{i{j}_{l}}$ is a random number between 2 and 5.

**Step 4.** Have cooperative enterprises interact using Equation (4).

**Step 5.** After setting the parameters in Step 1, repeat Steps 2–4 if per cycle t = t + 1, until t = T.