An awareness based approach to sustainability
After the upgrade of awareness of each agent , if the agent ’ s awareness rises beyond a given threshold the system updates the agent state , i . e . sets the membership of the agent to a new type .
Each type of agent has a different consumption and such consumption is updated according to an individual reduction goal . The overall consumption is evaluated based on the individual consumption and on resource production on the part of evangelists .
Each agent ’ s own resource consumption depends on an overall reduction goal . For each agent type , the goal is computed as the difference between the previous tick resource consumption and the individual reduction goal that has to be reached with a given rate .
The only agent type able to produce resources ( in addition to consuming them ) is the evangelist . The overall resource production includes the sum of all resources produced by evangelists .
The individual reduction goal varies according to agent type . SAM4SN relates awareness change to the availability of specific functions of a smart metering system . The rationale is that availability of smart metering functions enables the agent to know its own consumption of the resource and to identify an individual reduction goal . For blind agents , it is independent from the availability of any facilitating conditions , because blind agents want to increase their consumption despite any evidence of the need to reduce it . When the function of comparison with neighbors is available , agents know the consumption of other agents and can set their own reduction goal based on the minimum consumption of other agents .
The reduction goal depends on the minimum known consumption and is given by the difference between the previous consumption of the agent and the reference consumption of another agent that has the minimal consumption . Such difference is multiplied by a green-competition index . The minimal consumption of agents is computed as the consumption of the green agent whose consumption is minimal in the whole system . In a real situation , green agents accept to disclose their individual consumption in exchange for social reputation . The green-competition-index gives a weight of the aptitude of an agent to emulate the less consuming agents . The green competition index is small for lowaware agents and increases for high-aware agents until reaching the value of 1 for evangelists .
Agent ’ s awareness is also modified by the mechanism of social reinforcement . Each run a comparison takes place between the individual agent ’ s consumption trend and the global trend of resource use and when they are concordant , there is positive social reinforcement and such social reinforcement is added to the agent ’ s awareness , increasing it . In other words , reinforcement relies on the comparison between the global trend of resource use and the individual trend of consumption .
Our model identifies both individual consumption trend types ( i . e . reduction versus increment ) and an overall consumption trend . It is important to remark that agents know the global trend of the resource consumption , but not the overall reduction goal nor the global resource use level . When their behavior trends are concordant with the general consumption trend , the agents can “ reinforce ” their beliefs and this social reinforcement , in turn , changes their awareness . Positive reinforcement happens when both individual both global consumption trends are of reductions ; negative reinforcement happens when both individual both global
56