Optimization of Energy Consumption via Artificial ...€¦ · S-JPSET : Vol. 8, Issue 1, ISSN : 2229-7111 (Print) & ISSN : 2454-5767 (Online) copyright samriddhi, 2010-2016 Optimization

26S-JPSET : Vol. 8, Issue 1, ISSN : 2229-7111 (Print) & ISSN : 2454-5767 (Online) copyright samriddhi, 2010-2016

Optimization of Energy Consumption via Artificial Intelligence:A Study Sakshi Jaiswal*1, Awdhesh Gupta2 and Shivam Kumar Kanojiya3

1.* Student-B.Tech-IV Year, Department of Computer Science and Engineering, SMS, Lucknow, UP, India. e-mail : [email protected]. Department of Computer Science and Engineering, SMS, Lucknow, UP, India. e-mail : [email protected]. Department of Computer Science and Engineering, SMS, Lucknow, UP, India. e-mail : [email protected]

1. INTRODUCTION

The developments of global economic andscience led to the situation of environmentdeterioration and energy tense.

It is claimed that artificial intelligence is playingan increasing role in the research of managementscience and operational research areas. Intelligenceis commonly considered as the ability to collectknowledge and reason about knowledge to solvecomplex problems. In the near Future intelligentmachines will replace human capabilities in manyareas. Artificial intelligence is the study anddevelopments of intelligent machines and softwarethat can reason, learn, gather knowledge,communicate, manipulate and perceive the objects.

John McCarthy coined the term in 1956 as branchof computer science concerned with makingcomputers behave like humans. Efficient energyuse, sometimes simply called energy efficiency, isthe goal to reduce the amount of energy requiredto provide products and services. for instance,Installing fluorescent lights, LED lights or naturalskylights reduces the amount of energy requiredto attain the same level of illumination comparedwith using traditional incandescent light bulbs.Compact fluorescent lights use one-third theenergy of incandescent lights and may last from 6to 10 times longer. These are the certain benefitsof energy efficiency:- Lowering household energy bills - Energy

efficiency is the easiest, most affordable and

Abstract

In the future intelligent machines will replace or enhance human capabilitiesin many areas. Artificial Intelligence is the intelligence exhibited by machinesor software.”John McCarthy” who coined the tem in 1955 defines it as the“science and engineering of making intelligent machines”. As we are aware ofthe fact that energy consumption has grown tremendously over a few decadesin all over the world which is environmentally unfriendly. It is essential at thisstage of development to pause and critically examine the state of affairs via theapplication of AI in energy conservation and environmental system engineering.

In this paper we describe in a general way on how the existing applications ofAI techniques provide intelligent solution to optimize the energy conservationnow and in the future. Also, how Wireless Sensor and Actuator Networks areused to remotely monitor and control the environment according to the decisionsmade by the centralized reasoner and EEMS(Energy Efficiency ManagementSystem)provides effective energy saving measures and high quality energyconservation services Energy efficiency has nowadays become one of the mostchallenging task for both academic and commercial organizations and thishas boosted research on novel fields.

Publication Info

Article history :

Received : 29th Jan. 2016Accepted : 24th March 2016DOI : 10.18090/samriddhi.v8i1.11409

Keywords :Energy Efficiency Management; case-based reasoning; multi-agent system;artificial intelligence

*Corresponding author :Sakshi Jaiswale-mail : [email protected]

27copyright samriddhi, 2010-2016 S-JPSET : Vol. 8, Issue 1, ISSN : 2229-7111 (Print) & ISSN : 2454-5767 (Online)

27

Sakshi Jaiswal, Awdhesh Gupta and Shivam Kumar Kanojiya

most effective way for families to use energymore wisely and save money on both householdexpenses and transportation costs.

Improving business competitiveness - Energycosts affect a business’s bottom line. Businessesthat control their energy consumption enjoylower heating, electricity and transportationcosts.

Increasing energy available for export - Byusing energy more wisely, energy exporters likeNewfoundland and Labrador will haveadditional power to sell into global markets andthe resulting revenue can be invested in ourschools, hospitals and infrastructure.

Increasing comfort- Energy efficiency offersan opportunity to reduce energy costs, whileenhancing comfort.

Reducing local air pollutants - Energyefficiency can reduce the amount of local airpollutants that can come from sources like oilor wood.

This paper analyzes the optimization of energyvia artificial intelligence through AI based systemslike EEMS(Energy Efficency ManagementSystems) and how the combination of CBR (Case-Based Reasoning, CBR) and MAS (Multi-agentSystem, MAS) method is proposed to solve theexisting problems in energy efficiencymanagement. This paper introduces AI (ArtificialIntelligence, AI) to the design which will providevarious solutions to intelligent energyconsumptions now and in the future.

Energy efficiency management system (EEMS)provides effective energy saving measures andhigh quality energy conservation services; it willachieve the goal of energy conservation throughsystem integration, concordance and optimizationto the existing energy, which is particularlyimportant in today’s situation.The existing EEMSis based on ZigBee, cloud computing and business

intelligence technology, when making strategies,they don’t pay much attention on users’ feedback.For this reason, we present a novel approach whichcombines CBR with MAS to solve the existingproblems in EEMS. At the same time, users’feedback is added into the design, making thesystem more humanized and interaction with usersmore benign. Case Based Reasoning is the processof solving new problems based on the solutionsof similar past problems. It is not only a powerfulmethod for computer reasoning, but also apervasive behaviour in everyday human problemsolving. On the other hand Multi Agent System isa computerized system composed of multipleinteracting intelligent agents within anenvironment.MAS can be used to solve problemsthat are difficult or impossible for an individualagent to solve.

2. ARCHITECTURE OF EEMS

Decision-making is the kernel of EEMS, dueto the diversity of users’ requirements anduncertainties in using electricity, it’s usuallydifficult to find fixed rules or algorithms to supportdecision-making. Since CBR can use the oldexperience to solve new problems, providingflexible solutions, and compared with thealgorithms of quantitative, qualitative analysis ofexperience is easier for us to acquire, CBR issuitable in the decision-making link. But in fact,there is not much experience or similar cases inEEMS, it is hard for us to get a variety of modelsin processing problems, so we present MAS in thispaper to integrate different processing modes invarious industries and systems. The method ofMAS can enrich knowledge base of CBR, offeringplenty of source cases.

To sum up, this paper adopts the approach ofthe combination of CBR and MAS technology inEEMS, providing more diverse solutions indecision-making procedures. Systematic design isshown in figure 1.


Optimization of Energy Consumption via Artificial Intelligence: A Study

Energy Monitoring Agent: This part isresponsible for monitoring the usage ofelectricity. Energy monitoring agent comparescurrent energy usage with historical information,records the results and reports an emergencywhen abnormal data appears.

Energy Efficiency Analysis Agent: This agentis in charge of data analyzing. energy efficiencyanalysis agent will classify the characteristicsof different users, and finally find out theprinciple of energy utilization, which is helpfulto making effective decisions.

Decision-making Agent: This agent considersthe results of energy efficiency analysis agentand the present strategy completely, and makescorrect decisions when required. At the sametime, it will take the CBR output as an importantreference. Finally, the agent generates new andreasonable electricity scheme to guide users.

Energy Efficiency Diagnosis Agent: Analyzesenergy-using equipment from the system side,estimates current energy consumption and givessuplementary references to decision-makingagent, helping to improve energy efficiency.

User Feedback Agent: Uses the servicecondition to estimate the effectiveness of thesystem, evaluate the efficiency of the modelfrom the user side, and output auxiliarysuggestions to make real-time adjustments,improving decision-making plan continuously.

Information Maintenance Agent: This agentis in charge of all the system information’smaintenance and classification in a regular time,including user profile, energy-using equipment

information, energy utilization data and datafrom measuring points.

CBR Agent and CBR Knowledge AcquisitionAgent cooperate with each other to work fordecision-making agent. Some details will beintroduced in the following passage.

3. CBRCase-based reasoning (CBR) is put forward by

professor Schank in 1982. Case-based reasoning(CBR) has emerged as a major research area withinartificial intelligence research over the last decadedue to both its widespread usage by humans andits appeal as a methodology for building intelligentsystems. Conventional CBR systems have beenlargely designed as automated problem solvers forproducing a solution to a given problem byadapting the solution to a similar, previouslysolved problem.

Case-based reasoning is also applicable whenthe cases are more complicated, for example, whenthey are legal cases or previous solutions toplanning problems. In this scenario, the cases canbe carefully chosen and edited to be useful. Case-based reasoning can be seen as a cycle of thefollowing four tasks.Retrieve: Given a new case, retrieve similar casesfrom the case base.Reuse: Adapt the retrieved cases to fit to the newcase.Revise: Evaluate the solution and revise it basedon how well it works.Retain: Decide whether to retain this new case inthe case base.The process is shown in Fig.2

Fig.1 : EEMS Based on CBR/MAS

Fig.2: Case-Based Reasoning Progam System


29

3.1 CBR Knowledge Acquisition AgentThe internal structure CBR knowledge

acquisition agent uses three-layer architecture, andthe framework is shown in figure 3. Pattern Miningagents are located at the bottom layer and areresponsible for gathering and analyzing models invarious industries. Model integrating agents arelocated at the middle layer and are responsible forgathering the output of pattern mining agents,forming a large scale of data for data analysis anddecision-making. Knowledge acquisition agent islocated at the top layer and is in charge of neateningall the models provided by the other two layers,offering problem-solving models as many aspossible.

3.2 CBR AgentCase-based reasoning has been formalized for

purposes of computer reasoning as a four-stepprocess:[1]

1. CBR Retrieve Agent: Given a target problem,retrieve from memory cases relevant to solvingit. A case consists of a problem, its solution,and, typically, annotations about how thesolution was derived.

2. CBR Reuse Agent: Map the solution from theprevious case to the target problem. This mayinvolve adapting the solution as needed to fitthe new situation

3. CBR Revise Agent: Having mapped theprevious solution to the target situation, test thenew solution in the real world (or a simulation)and, if necessary, revise

4. CBR Retain Agent: After the solution has beensuccessfully adapted to the target problem, storethe resulting experience as a new case inmemory.

4. MULTI AGENT SYSTEM (MAS)

A multi agent system is a computerized systemcomposed of multiple interacting intelligent agentswithin an environment .Multi agent systems canbe used to solve problems that are difficult orimpossible for an individual agent to solve.Intelligence may include some methodic,functional, procedural approach, algorithmicsearch or reinforcement learning. It consists ofagents and their environment. They may be: Passive agents or agents without goals. Active agents with simple goals. Cognitive agents which contain complex

calculations.Agents environment can be divided into : Virtual Environment Discrete Environment Continuous Environment

The agents in MAS have several importantcharacteristics:


Fig.3: Knowledge Acquisition Agent of CBR

Fig.4 : Coordination of CBR Agent and Decision-MakingAgent


Autonomy: The agents are atleast partiallyindependent, self aware, autonomous.

Local views: No agent has a full global viewof the system, or the system is too complex foran agent to make practical use of suchknowledge.

Decentralization: There is no designatedcontrolling agent.

4.1 Case Based Reasoning Method Apply onMulti-Agent Systems

Multi-Agent System technique is employed inCBR system for the purpose of retrieving, reusing,adapting the cases in CBR System. This sectionexplores a framework that integrates the multi-agent and case-based reasoning techniques tosupport the dynamic and problem-orientedknowledge sharing among supply chain members.The framework is characterized by differentialknowledge sharing levels depending upon theapplications as well as the knowledge creation andreuse based on the previous knowledge in theproblem area. It also provides a new tool for thefield of inter-organizational knowledgemanagement. The CBR system retrieves casesrelevant to the present problem situation from thecase base and decides on the solution to the currentproblem on the basis of the outcomes fromprevious cases CBR System based on MA systemconsists of the following main agents:

Retriever Agent: When a new problem isentered into a case based system, a retrieverdecides on the features similar to the storedcases. Retrieval is done by using features of thenew cases as indexes into the case base.

Adapter Agent: An adapter examines thedifferences between these cases and the currentproblem .It then applies rules to modify the oldsolution to fit the new problem.

Refiner Agent: A refiner critiques the adaptedsolution against prior outcomes. One way to

do this is to compare it to similar solutions ofprior cases. If a known failure exists for aderived solution, the system then decideswhether the similarities are sufficient to suspectthat the new solution will fail.

Executer Agent: Once a solution is critiqued,an executer applies the refined solution to thecurrent problem.

Evaluator Agent: If the results are as expected,no further analysis is made, and the cases andits solution are stored or use in future problemsolving. If not, the solution is repaired.

4.2 User Feedback AgentThough providing efficient scheme in using

electricity, user’s feedback is ignored to somedegree, and the energy-saving effect is not thatobvious. In this paper we introduce MAS into thedesign to enhance users’ feedback and interactionwith the system.It consists of three layers: At the bottom layer Smart meters are present

and are responsible for gathering and savingusers’ information in using electricity.

Fig.5 : Multi-agent based CBR System



31

The middle layer includes potential assessmentagent, scheme optimization agent and usercomfort agent. Agents in middle layer have tocommunicate with the top agent frequently.

User feedback agent is at top layer and deploysthe functions of each layer. The ramework isshown in fig 6.

Potential Assessment Agent: Assess users’energy saving efficiency, and classify userproperty and application in using electricity.This agent monitors energy consumption inmany aspects continuously and calculates thepotential of each user.

Scheme Optimization Agent: This part isresponsible for predicting the trend in usingelectricity, and makes preparations for energyconservation.

User Comfort Agent: Collect users’ habits andemergency in power utilization through users’feedback, and compare them with the schemegenerated by the system. All the informationshould be transmitted to the decision-makingagent.

5. MAS COMMUNICATION

There are two kinds of communication modesof MAS, and they are synchronous messaging andasynchronous messaging. In user feedback agentwe adopt the communication mechanism ofsynchronous messaging. The blackboard is apublic area for changing information andknowledge. Smart agents regard the blackboardas a base for sharing information, and they don’t

have communications directly between each other,and their duties are finishing their own taskseparately. There are three layers in user feedbackagent, the bottom layer share the information theygathered on the blackboard and the middle layeragents can evaluate themselves or optimize tasks.At the same time agents in one layer can also havemutual learning through the blackboard. Finally,the top layer agent will find optimal solutions fromthe blackboard, providing useful references fordecision-making agent. For the dynamiccollaboration tasks solving system, the number ofthe board may be changed along with the changeof situations and needs.

6. CONCLUSION

This paper presents the approach of combiningtwo artificial intelligence methods in EEMS i.e.CBR and MAS technology, providing morediverse solutions in decision-making process. Inaddition, we regard the user comfort as animportant aspect in the design, users’ feedback canaffect the decision-making link to a certain extent,providing users with effective yet comfortableservices, and reaching the goal of saving energyat the same time.

7. BIBLIOGRAPHY Paper [1]and[ 2] is about the application of

energy efficiency management technology in acommercial and industrial areas, paper [3]introduce the technology of energy efficiencymanagement in intelligent villages, paper [4, 5]mention the application of managementtechnology of energy efficiency in intelligentbuildings, comparision of policies of energyefficiency management in China and in othercountries in paper [6, 7] which provides new ideasfor energy saving and emission reduction. Theexisting EEMS combines ZigBee, cloudcomputing and business intelligence technology.Paper [8] introduce the method of three layers

Fig.6 : User Feedback Agent



architecture, realizing the collect ion andmanagement of electric equipment’s information.improvement of intelligence of EEMS Expertsystem is described in paper [9]. Paper [10]provides economic and efficient utilization ofEEMS for users and proposes ladder electricityprices. paper [11] introduces EEMS design andimplementation. Paper [12] proposes a cloudcomputing data centre resources energy efficiencymanagement framework based on IaaS model.Paper [13] introduces the approach of “ThinkHome project”, through which optimizing ofenergy efficiency and user comfort at the same timecan be achieved, yet always acknowledging theresidents’ desires.Wireless Sensor Networks areused to collect data and it also controls the EEMSin a distributed way in Paper[14] which explaintest-bed. Paper [15] presents EECLOUD an energyefficiency management method; the experimentalresults show that this method achieves greaterenergy savings over existing methods. Paper [16]proposes an EEMS for the use of the powergenerated by a Smart Home and the powerconsumed by the building’s electric appliances.Paper [17] presents a decision support system,based on intelligent energy storage, which is ableto manage both the electric power and the smarthome devices of a house in order to optimize thelocal energy consumption. Paper[17] presents theenergy efficiency management system based onCBR/MAS technology to optimize the energyconsumption via artificial intelligence.

REFERENCES[1] CHEN Hua-Lin, YAO Jian-gang, HUANG Shi-wen.

Research of energy efficiency assessment model inpower demand side[J]. POWER DSM, 2010(4) (inChinese).

[2] HUANG Qiang. Energy Efficiency ManagementInformation System and Application [J]. POWERSUPPLY, 2010(2) ( in Chinese).

[3] Luhua Z. The Design and Implementation of FamilyComprehensive Energy Management System Facingthe Smart Power [J]. Electrical Measurement &Instrumentation, 2010(537)(in Chinese).

[4] Binruo Z. Design and Implementation of IntelligentBuilding Electricity Energy Management System [J].East China Electric Power, 2010(4)(in Chinese).

[5] Wei L. Study and Application of EEMS forIntelligent Buildings [J]. POWER SUPPLY, 2011(3)[6] Shihai C. Measures of promotion the actions andpolicies of energy efficiency management [J].EXPERT VIEWS, 2009(4)(in Chinese).

[6] Xin H. Europe and the United States national energyconservation policy evolution trend and theenlightenment to China[J]. Economic Aspect,2008(9)(in Chinese).

[7] Bingzhen Z. Application of energy efficiencymanagement technology toward residential user[J].ELECTRIC POWER IT, 2011(4) (in Chinese).

[8] Pengfei L. Application Research of EEMS in theIntelligent Building Based on Expert System [J].Jiangsu Construction, 2013(158) (in Chinese).

[9] Jiye W. Design of intelligent home EEMS based onthe intelligent electricity interactive service platform[J]. ELECTRIC POWER IT, 2012(12) (in Chinese).

[10] Aimei Y, Yingxin X, Lifang Y. Research andApplication of EEMS for Intelligent Park[J]. ElectricPower Information, 2014(7) (in Chinese).

[11] Wei H. Energy efficiency Management ResearchBased on Genetic Algorithm in Cloud Computing[D]. Wuhan University of Technology, 2013(inChinese).

[12] Kastner, W. ; Kofler, M.J. ; Reinisch. Using AI torealize energy efficient yet comfortable smarthomes[J]. Factory Communication Systems,2010(18).

[13] Ahmadi, S.A. ; Shames, I. ; Scotton. Towards moreEfficient Building Energy Management Systems[J].Knowledge, Information and Creativity SupportSystems, 2012(21).



33

[14] Yanfei Li ; Ying Wang ; Bo Yin ; Lu Guan. An EnergyEfficient Resource Management Method inVirtualized Cloud Environment[J]. NetworkOperations and Management Symposium, 2012(14).

[15] Nazabal, J.A;Fernandez-Valdivielso, C. ; Falcone.Energy Management System Proposal for EfficientSmart Homes[J]. New Concepts in Smart Cities:Fostering Public and Private Alliances, 2013.

[16] BotonFernandez, V.; Lozano-Tello, A. ; PerezRomero. Intelligent Decision Support System for theEfficient Energy Management in HouseholdEnvironments [J].Information Systems andTechnologies, 2013.

[17] K.Y. Xu, D.W. Zhou ;Y.Q.Chen; W.B.Lu; Researchon Intelligent Energy Efficient Management Systembased on CBR/MAS.International SeminarConference on Computer Information Systems andIndustrial Applications (CISIA 2015)


Documents

Optimization of Energy Consumption via Artificial ...€¦ · S-JPSET : Vol. 8, Issue 1, ISSN : 2229-7111 (Print) & ISSN : 2454-5767 (Online) copyright samriddhi, 2010-2016 Optimization