4.4 Dynamic Supply Control and Stabilization Mechanismage 2

The SCG ecosystem adopts an SDA (Supply Decision Algorithm) that dynamically adjusts the supply of tokens and NFTs based on real-time indicators such as market liquidity, user participation, NFT circulation volume, and transaction density. This algorithm is integrated with Solana’s high-performance on-chain infrastructure and aims to maintain economic stability and a balance of scarcity through a distributed reinforcement learning policy structure.

1. SDA Layered Structure and Computation Engine Composition

SDA consists of the following computational layers:

• Time Series Analysis Engine (LSTM-based): Analyzes time-series inputs such as NFT transaction volume, token velocity, and price volatility to forecast medium-term supply-demand trends.

• State Classification Engine (GMM-based): Probabilistically classifies market conditions and branches into optimal supply policy groups based on conditions.

• Policy Learning Engine (DQN-based): Evaluates policies for reward maximization and selects supply policies based on Q-functions.

These three modules are integrated into the internal pipeline of the SDA, where reward function updates and policy improvements are iteratively performed based on changes in supply volume.

1. Formula-Based Supply Adjustment Logic

SDA mathematically defines a loss function and a reward function for policy selection and proceeds with optimization under the following structure:

A. Loss Function Definition:

This loss function evaluates the sensitivity of supply adjustment and models the impact of supply fluctuations on the market.

B. Reward Function Definition:

The DQN agent updates the Q-function along with an experience memory and continues learning to select optimal policies.

1. On-Chain Triggers and Execution Conditions

• Automatic Execution Conditions: Supply adjustment is triggered on-chain when the following conditions are met:

- Issuance volume surge of more than 30% compared to average trading volume

- 10-day average NFT non-usage rate exceeding 65%

- Occurrence of DAO voting participation decline and liquidity shrinkage indicators

• On-Chain Execution Method:

- Supply Expansion: Increase in reward pools, enhancement of NFT upgrade probabilities, and automatic issuance of bonus tokens

- Supply Contraction: Token burning, limitation on NFT issuance, and contraction of reward pools

All of these triggers are executed automatically through conditional statements and event-handling functions in Solana smart contracts, and all state transitions are recorded and verifiable.

1. Post-Execution Feedback Loop

After the SDA policy is executed on-chain, the following indicators are tracked to update the Q-function and policy:

• SCG price volatility reduction rate after 7 days

• Circulation rate of distributed tokens and NFT turnover

• Reward claim rate and changes in DAO participation

• Changes in liquidity balance in the market (based on Solana DEX)

This feedback is stored in the experience replay memory of the SDA and serves as the basis for learning new supply policy selections.

SDA serves as the core autonomous regulation system for distribution stability in SCG. It unifies quantitative supply judgment and policy execution into a single reinforcement decision-making framework. Through this, the SCG ecosystem establishes a technological infrastructure that manages demand reflection, scarcity maintenance, and price fluctuation suppression within one computational model—while continuously securing compatibility and scalability with the Solana blockchain.