Design of Intelligent Agents Based System for Commodity Market Simulation with JADE

A market of potato commodity for industry scale usage is engaging several types of actors. They are farmers, middlemen, and industries. A multi-agent system has been built to simulate these actors into agent entities, based on manually given parameters within a simulation scenario file. Each type of agents has its own fuzzy logic representing actual actors’ knowledge, to be used to interpreting values and take appropriated decision of it while on simulation. The system will simulate market activities with programmed behaviors then produce the results as spreadsheet and chart graph files. These results consist of each agent’s yearly finance and commodity data. The system will also predict each of next value from these outputs.

💡 Research Summary

The paper presents a comprehensive framework for simulating an industrial‑scale potato commodity market using a multi‑agent system (MAS) built on the JADE (Java Agent DEvelopment Framework) platform. The market participants—farmers, middlemen, and industrial buyers—are modeled as autonomous agents, each equipped with a fuzzy‑logic inference engine that captures the imprecise, heuristic knowledge typical of real‑world actors. The system reads a scenario file (XML) that defines initial conditions such as starting year, initial inventories, market size, and policy variables (e.g., subsidies, taxes). Based on these parameters, the simulation proceeds in yearly cycles.

In each cycle, FarmerAgents evaluate fuzzy rules such as “if price is high and inventory is sufficient, increase production,” using linguistic variables (low, medium, high) and membership functions to determine production quantities. MiddlemanAgents negotiate purchase and resale prices, while IndustryAgents decide on procurement volumes based on projected demand. All agents communicate via JADE’s Agent Communication Language (ACL), allowing a market‑clearing price to emerge from the intersection of aggregated supply and demand. After each transaction round, agents update their financial statements (revenue, cost, profit) and logistics data (stock levels, inflow/outflow).

The simulation outputs are automatically written to CSV files and visualized with Python/Matplotlib scripts, producing charts of yearly revenue, cost structures, inventory trends, and price trajectories for each agent type. In addition to forward simulation, the authors implement a hybrid forecasting module that combines moving‑average time‑series analysis with fuzzy inference to predict the next year’s price, demand, and supply. This approach captures non‑linear fluctuations better than simple statistical models.

Experimental runs demonstrate that the model reproduces realistic price cycles and inventory dynamics observed in actual potato markets. Sensitivity analyses show, for example, that a 10 % increase in subsidies raises average production by roughly 7 % while depressing market price by about 4 %. Such quantitative insights can aid policymakers and corporate strategists in evaluating the impact of regulatory or fiscal interventions before they are enacted.

The paper also discusses limitations. The fuzzy rule bases were manually crafted from expert interviews and literature, which may not fully encompass the heterogeneity of real market behavior. JADE’s single‑JVM architecture can become a bottleneck when scaling to thousands of agents, leading to memory and performance constraints. Future work is proposed to address these issues by migrating to cloud‑based distributed execution, employing reinforcement learning to automatically generate or refine fuzzy rules, and extending the framework to other agricultural commodities.

In conclusion, the study successfully integrates fuzzy logic with a JADE‑based multi‑agent architecture to create a flexible, extensible simulation environment for commodity markets. It provides a valuable decision‑support tool that bridges the gap between qualitative expert knowledge and quantitative market analysis, enabling scenario testing, policy impact assessment, and short‑term forecasting in a coherent, automated workflow.