US20110106327A1

Movatterモバイル変換

Info

Publication number: US20110106327A1
Application number: US12/640,020
Authority: US
Inventors: Zhi Zhou; Juan Manuel De Bedout; John Michael Kern; Emrah Biyik; Ramu Sharat Chandra
Original assignee: General Electric Co
Current assignee: General Electric Co
Priority date: 2009-11-05
Filing date: 2009-12-17
Publication date: 2011-05-05
Also published as: US8359124B2; US20110106328A1

Abstract

A method for optimizing customer utility usage in a utility network of customer sites, each having one or more utility devices, where customer site information is communicated between each of the customer sites and an optimization server having software for optimizing customer utility usage over one or more networks, including private and public networks. A customer site model for each of the customer sites is generated based upon the customer site information, and the customer utility usage is optimized based upon the customer site information and the customer site model. The optimization server can be hosted by an external source or within the customer site. In addition, the optimization processing can be partitioned between the customer site and an external source.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the benefit of pending provisional application Ser. No. 61/258,412, filed Nov. 5, 2009. This application is related to co-pending application entitled “ENERGY OPTIMIZATION SYSTEM” filed concurrently.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

This invention was made with Government support under contract number DE-FC26-06NT42847 awarded by U.S. Department of Energy. The Government has certain rights in the invention.

BACKGROUND

Efficient energy management systems are desired by utilities and customers in order to reduce costs and peak demand to provide more efficient energy use. Various measures have been developed to address energy management goals. For example, energy efficient devices and appliances have been developed to enable customers to manage energy consumption. In addition, utilities have provided some customers with direct control for certain electrical loads, such as heating or cooling devices, pool pumps, and so on, to allow utilities to control energy and power usage. Also, utilities have implemented demand response programs for residential as well as industrial customers in order to provide reduced peak demand requirements and to effectively manage energy consumption.

While a number of developments have been made to enable customers and the utilities to manage consumption on the grid, the integration, automation, and optimization of these devices into an intelligent system that monitors and manages consumption and effectively communicates information between the utilities and the customers are still being developed.

For these and other reasons, there is a need for the present invention.

SUMMARY

A method for optimizing customer utility usage in a utility network of customer sites, each having one or more utility devices, where customer site is communicated between each of the customer sites and an optimization server having software for optimizing customer utility usage over one or more networks, including private and public networks. A customer site model for each of the customer sites is generated based upon the customer site information, and the customer utility usage is optimized based upon the customer site information and the customer site model. The optimization server can be hosted by an external source or within the customer site. In addition, the optimization processing can be partitioned between the customer site and an external source.

BRIEF DESCRIPTION OF THE DRAWINGS

The nature and various additional features of the invention will appear more fully upon consideration of the illustrative embodiments of the invention which are schematically set forth in the figures. Like reference numerals represent corresponding parts.

FIG. 1 illustrates a utility management system according to an exemplary embodiment of the invention;

FIG. 2 illustrates a utility management system with a centralized optimization according to an exemplary embodiment of the invention;

FIG. 3 illustrates a utility management system with another centralized optimization according to an exemplary embodiment of the invention;

FIG. 4 illustrates a utility management system with a distributed optimization according to an exemplary embodiment of the invention;

FIG. 5 illustrates a utility management system with a hybrid optimization according to an exemplary embodiment of the invention;

FIG. 6 illustrates an information flow diagram according to an exemplary embodiment of the invention;

FIG. 7 is a flow diagram of an optimization process according to an exemplary embodiment of the invention;

FIG. 8 is a flow diagram of a system/model identification process according to an exemplary embodiment of the invention;

FIG. 9 is a flow diagram of a real-time utility consumption and cost estimate process according to an exemplary embodiment of the invention;

FIG. 10 is a screen shot of a user interface according to an exemplary embodiment of the invention; and

FIG. 11 is a flow diagram of a direct load control process according to an exemplary embodiment of the invention.

While the above-identified drawing figures set forth alternative embodiments, other embodiments of the present invention are also contemplated, as noted in the discussion. In all cases, this disclosure presents illustrated embodiments of the present invention by way of representation and not limitation. Numerous other modifications and embodiments can be devised by those skilled in the art which fall within the scope and spirit of the principles of this invention.

DETAILED DESCRIPTION

As used herein, the term “module” refers to software, hardware, firmware, or any combination of these, or any system, process, or functionality that performs or facilitates the processes described herein.

Embodiments of the invention are described with reference to a particular example of an electric utility. However, this example is only used to facilitate description of embodiments of the invention and in no way limits the scope of the invention. Embodiments of the invention can be applied to any utility network or applied to any other suitable application. In addition, the customer sites disclosed herein may be residential homes, apartment buildings, industrial, commercial, educational, military, and/or office buildings, for example.

According to an embodiment of the invention, an Energy Optimization System (EOS) module includes an energy management platform that interfaces with smart meters, smart loads and devices, and sensors, within a customer site such as a customer home or commercial building as well as with smart energy generation and energy storage components. The customer interfaces with the EOS module through a user interface having a display. The EOS module is an energy optimizing engine, which is based on model predictive controls. The energy optimizing engine utilises dynamic variable data such as forecast weather, solar insolation, and cost of electricity, along with historic and real-time data from sensors within the customer site and a model of the customer site to plan the optimal usage of the energy generated at the customer site, the storage level of energy storage and the operation of the loads. In addition, according to embodiments of the invention, the customer is provided with a real-time estimate of energy uses and monthly savings, as the customer makes adjustments to his/her preferences on the user interface. The utility is provided with the validation and quantification of actions from demand response events and the estimation of customer actions in the event of a utility demand response request.

The energy optimizing system module according to embodiments of the invention can be utilized in many different conservation or incentive programs and scenarios. It can be used for many different functions such as energy site automation, education and information, and tracking and reporting. The EOS module enables the customer and the utility to capture a significant entitlement of smart grid ownership and load ownership as it allows the various subsystems to work in unison as a single system.

With regard to site energy optimization and automation functions, the EOS module optimizes and automates the energy savings, empty site setbacks, pre-cooling/pre-heating of the site for peak load reduction, and peak load reduction operation of all the major loads including appliances, heating, ventilating and air conditioning (HVAC), water heating, lighting, and miscellaneous electric loads. The EOS module receives pricing signals from a utility smart meter, such as time-of-use rate tariff, real-time pricing tariff or critical peak pricing, and determines optimized energy usage based on the received information. The EOS module also receives forecast weather and solar insolation from weather servers, which reside on a public network such as the Internet, and parses it for usage in the optimizer module. The EOS module interfaces with all major loads via a wireless network using an interface such as the Zigbee protocol, for example. Communication between the customer sites and the utility occurs via a public network or a WAN such as the Internet. Customer specific or sensitive information can be communicated via a secure private network, which may be a wireless network.

With regard to the education and informational function, the EOS module provides real-time estimates on the customer display of energy usage and utility cost. The estimates can be showed for what is saved at the end of the month or for the next billing cycle (e.g., 30 days from now). The EOS module utilises historical trends and forecast information to make this estimate.

With regard to tracking and reporting, the EOS module reports on the customer display real-time usage of electricity. The information is stored and is trended over various time intervals (e.g., in days, weeks, months, and years). A similar tracking and reporting scheme can be easily implemented for water usage, natural gas usage as well as solar or any other distributed power and energy production.

According to embodiments of the present invention, the EOS module uses a set of mathematical models to optimize and control the energy system. The models allow the energy optimizer to predict how the customer site reacts to stimuli such as weather, HVAC operation, and occupancy, for example. According to this embodiment, the models are constructed and updated automatically by the EOS module using a system identification (ID) process to enable self-calibrating/updating of the models with minimum customer interaction/intervention. The models can also be used to track changes to the site structure over time as a means for monitoring and diagnostics.

Since the EOS module relies on a set of mathematical models that represent the physics/dynamics of the site, the system can have the capability/ability to self-identify, construct, and update periodically the models by initiating a set of runs that are non-invasive/non-intrusive to the customers. The system ID is initiated by the EOS module and/or at a time of any major site modifications that would significantly impact on the site. In addition, the system ID process can adapt to dynamic variables such as seasonal climate changes, for example, with fine-tuning of the models by self-initiating a subset of the system ID runs. The system ID algorithms detect abnormal conditions, such as open windows and doors, to prevent the system ID from running. Such conditions would, otherwise, skew the system ID and the mathematical models.

According to embodiments of the invention, the energy management system includes the EOS module, connectivity to the utility, connectivity to loads and/or resources at customer sites, and a user interface or human machine interface. In addition, various configurations of the system are possible, such as a centralized architecture, a distributed architecture, and a hybrid architecture, for example.

FIGS. 1-3 show anenergy management system100 according to an exemplary embodiment of the invention where the EOS module is arranged outside of the customer site on a centralized server. According to this embodiment, the energy optimizer module monitors and optimizes utility usage at a number of customer sites. Referring toFIG. 1, theenergy management system100 includes at least oneutility control center102 and customer sites104-104(n). Eachcustomer site104 includes anenergy manager112 having a processor114, a memory116, a user interface118 and adisplay120. The user interface118 can include a keyboard or touch screen, for example. The processor114 runs programs for monitoring and controlling the operation of various utility devices such assensors122,renewable resources124,storage devices126, plug in electric vehicles (PEV) or plug in hybrid electric vehicles (PHEV)128, and other appliances and loads130. Thesensors122 can include thermostats, occupancy sensors, humidity gauges, and other suitable devices. The renewable resources can include solar and/or wind power devices, for example. Theenergy manager112 communicates with the various components using any of a number of interfaces or protocols including Zigbee, Z-Wave, Wi-Fi, and communication over power lines or PLC, for example. Theenergy manager112 further supports communication via an advanced metering infrastructure (AMI)110.

Theenergy manager112 at each of the customer sites exchanges information with the utility control center over a

public network

106,106aand/or aprivate network108. The public network is a WAN such as the Internet. The networking protocols for communication according to embodiments of the invention include Wi-MAX, Broadband over Power Lines (BPL), PLC, DSL/Broadband, and any other suitable standard and/or private protocols. Customer specific or sensitive information can be exchanged via theprivate network108, while other communication occurs over the

Internet

106,106a. This connectivity provides various types of communication to support anticipated market growth. Three communication type examples include public pricing signaling, consumer specific signaling, and control signaling. Public pricing is the communication of material, which is publicly available. Consumer-specific signaling would be the signaling that supports a home energy management system. Control signaling are those signals used to support load shedding. Each signal type warrants individual security and privacy analysis and treatment.

Although the optimization algorithms are located outside of the customer site in the embodiments shown inFIGS. 1-3, the end-users can still have full access and control to the EOS module via a customer account on the server and input preferences, overwrites, or bypasses, for example, via acustomer portal131.Customers104 as well as theutility control center102 have access to thecustomer portal131 via thepublic network106 and/or theprivate network108. Thecustomer portal131 provides interfaces that allow customers to monitor and control utility consumption at thecustomer site104. Thecustomer portal131 can also include online tools and utility applications.

As shown inFIG. 2, theutility control center102 includes theEOS module132. In addition, theutility control center102 includes an energy management system (EMS)module134 that performs load forecasting for the network, and monitors, controls, and optimizes the performance of the generation and transmission system. A Supervisory Control And Data Acquisition (SCADA) module136 provides real time information at different points in the grid and also provides local controls. An Outage Management System (OMS)module138 monitors load status information and outage restoration information for thecustomer sites104 in the network. Some of the functions performed by theOMS module138 include failure prediction, providing information on the extent of outages and impact to customers, and prioritizing restoration efforts. TheOMS module138 operates based on a detailed network model of the distribution system that is generated and maintained by a Geographic Information Systems (GIS)module140. A Distribution Management System (DMS) module142 provides real-time response to adverse or unstable network conditions by providing information on load status and load response. The DMS module142 manages the response to alarms and/or events. Customer information including service contract information, participation in incentive and/or demand response programs, and contract price information, for example, is monitored and controlled by the Customer Information System (CIS)module144. A Direct Load Control (DLC)module145 controls and manages customer site devices such as the thermostat—HVAC, water heater, pool pump, washer, dryer, dishwasher, LCD/Plasma TV, plug loads (e.g., computers, computer peripherals/accessories, fax machine, power supplies), refrigerator, and lighting, for example. These are mostly discrete types of devices that have on/off, eco-mode/normal mode, or multiple discrete power saving modes (e.g., dimmable lighting). Customer billing is performed by thebilling module146. A Demand Response (DR) module148 responds to network events. A Network Management System (NMS)module150 provides communication management and provisioning for theEOS module132 as well as the other modules and thecustomer sites104. TheEOS module132 utilizes information from the various modules in theutility control center102 and from theenergy manager112 at each of thecustomer sites104 to monitor and optimize utility usage at thecustomer sites104.

Theutility control center102 also includes adata storage unit152 for storing data such as historical data for each customer site in the network based on information from theEOS module132, theEMS module134, the DMS module142, the SCADA module136, and theOMS module138, for example. The historical data can include information on customer utility usage including load type, time of use (TOU), duration of use, and shed or demand response events, for example. The customer usage information stored in thedata storage unit152 can be updated periodically (e.g., hourly, daily) with load data including hourly load and hourly price over a twenty four hour period, environmental data including weather information (e.g., temperature, humidity, wind speed, and solar insolation) and date and time information such as day of the week, season, etc. Finally, theutility control center102 includes a user interface module154 and anoperator console156.

In the embodiment shown inFIG. 2, theEOS module132 is arranged within theutility control center102. However, embodiments of the invention are not limited in this regard. For example, as shown inFIG. 3, theEOS module132 can be hosted outside of theutility control center102 and can be hosted by a party other than the utility. It can also be integrated with other energy management systems, at the transmission and distribution levels, for example, within the entire utility network.

FIG. 4 shows another energy management system200 according to an embodiment of the present invention. According to this embodiment, theEOS module214 is arranged in theenergy manager212 at each of thecustomer sites104. Eachcustomer site104 has its own dedicatedenergy optimizer module214. In other embodiments, theEOS module214 can reside in a smart thermostat, a cable set-top box, a “smart” in-home display, or other computing/smart devices with sufficient computing capability and memory in order to perform optimization and having the ability to display parameters and accept user inputs. In contrast to the embodiment shown inFIGS. 1-3, the architecture in this embodiment has all of the energy management and optimization functions and intelligence installed on a computer or personal PC at thecustomer site104. TheEOS module214 communicates and interfaces with all of the other customer site components, such as meters and appliances, at the customer site. Communication with theutility control center102 and the devices at the customer site are performed in the same manner as that described above with respect toFIGS. 1-3.

FIG. 5 shows yet another embodiment of the invention. According to this embodiment, the energy optimization system module is split between the central server where theEOS module132 is arranged either at theutility control center102 or hosted outside of theutility control center102, and theenergy manager212 or other smart panel at the customer site to form a hybrid architecture. This hybrid approach of distributing energy management/optimization algorithms/intelligence between TheEOS module132 and theEOS module214 at thecustomer site104 offers more system flexibility. According to this embodiment, the split of the intelligence for the hybrid architecture is application dependent. One example of this hybrid approach of distributing energy management optimization algorithms and intelligence between theEOS module132 and theEOS module214 provides that the data pre-processing such as data filtering and averaging, for example, diagnostics and prognostics are performed by theEOS module214 at the customer site, and all other computing and memory intensive optimizations reside on the centralized server orEOS module132. The hybrid architecture according to this embodiment of the invention can best utilize all of the local and global (central) resources effectively to minimize the network data traffic and maximize the system efficiency and speed.

It is to be understood that the function of the

EOS modules

132 and214 can be distributed based on the needs of the application and that communication between the utility control center and thecustomer sites104 remains the same as inFIGS. 1-4.

In order to facilitate the description of the embodiments of the invention, the details for a singleutility control center102 and a singleutility customer site104 are shown in the Figures. However, it should be understood that embodiments of the invention are not limited to these numbers, and that the utility can have any number of utility control centers and customers. Also, in order to facilitate the description of embodiments of the invention below, reference will be made toEOS module132. However, it is to be understood, that the embodiments can includeEOS module214 alone or a combination of

EOS module

132 and214.

TheEOS module132 provides a system level optimization. According to embodiments of the invention, the optimization algorithms are coupled with an adaptive customer site model that is updated automatically via a system identification process to adapt to building structure changes (e.g., renovation and/or aging) and seasonal weather changes, etc. This advanced adaptive feature ensures consistent high performance of the optimization algorithms of theEOS module132 with maximized energy, power, and cost reduction. In addition, according to an embodiment of the invention, theEOS module132 provides a real time estimate of costing savings/benefit as a result of customer preference/setting changes. This provides real-time feedback to the customers on how much savings can be expected by changing, for example, thermostat temperature setpoint.

According to embodiments of the present invention, the EOS module or utility optimizer, implements utility usage efficiency beyond the open-loop energy management system data display and analysis and direct load control, to a closed-loop customer site automation/energy optimization to provide system level optimization that takes into account all of the resources available to optimize the energy utilization. According to an exemplary embodiment, theEOS module132 optimizes utility usage for customers by minimizing monthly utility bills such as electric, gas, and oil, for example, reducing wasteful utility utilization, and providing for demand-response participation. The objectives of each customer depend upon customer preferences input into the system via the user interface. In the exemplary embodiment, optimization is achieved by controlling multiple or single zone thermostat settings, battery charging or discharging, hot water heater operation, lighting operation, deferrable loads, sheddable loads, and Photo Voltaic (PV) inverter power flow. The EOS module monitors instantaneous aggregated power consumption from the utility meter, instantaneous circuit-specific power consumption through current sensors, PV power production, export of PV Watts when net-metering is available, zonal temperatures from zonal thermostats, utility cost for customers participating in a time-variable rate program, battery state of charge from battery charging unit, and export of PV VARs (Volt-Amp-Reactive), for example.

FIG. 6 illustrates the information flow for theEOS module132 for optimizing utility usage at customer sites.Input information600 and measured/observedinput information604 are used in a system identification (ID)process602 to generate and update a mathematical model of the customer site. Theinput information600 along with the customer site model from thesystem ID process602 is used to generateoptimization information603 for the customer site. In an exemplary embodiment, the input information includes user input, acquired and predicted input, system parameters and constraints, as well as control objectives.

The user input includes, for example, preferred setpoints, upper and lower bounds on a utility parameter such as zone and/or water temperature, and deferrable and sheddable load priority preferences. The acquired and predicted input includes, for example, ambient temperature forecast, solar insolation, cost of utility, and load demand. The system parameters and constraints can include the number of thermal zones, HVAC maximum power, battery charge/discharge rates and state of charge depth, and equipment cycling limits, for example. The control objectives include minimizing utility bill, reducing utility utilization at the customer site, and responding to utility demand signals, for example. The measured/observedinput604 includes room and/or water temperature, battery state of charge, and current consumption, for example. Of course, the input information and the measured/observed information can include information in addition to the examples noted herein in light of the specific application.

Optimization information

603 is generated using model predictive control (MPC) based on theinput information600, the systemID model estimate602, and the measured/observedinput604. The electrical and thermal loads, as well as renewable generators (if they are present) in a customer site are variable. When the predicted values of these quantities are known ahead of time, it is possible to incorporate them into an optimization strategy which minimizes the total, predicted cost of operation (e.g. over a time horizon of 24 hours, known as the “prediction horizon”). This predicted cost includes the cost of buying electric power/gas from the grid or utility, as well as the profit from selling power to the grid (if this is an option). The optimal solution can be recomputed periodically (e.g., every 15 minutes) in order to take advantage of updates of the predicted loads and prices. The optimization information is used to generate commands that are sent to the utility devices and/or resources at the customer site to optimize utility usage. According to embodiments of the invention, the input information, measured/observed input, system ID information and optimization information are updated periodically. TheEOS module132 leverages historical utility usage statistics compiled throughout its operation to establish estimates of periodic (e.g., hourly) energy consumption, ambient temperature, solar insolation and other variables that can be used for planning how to best operate and control the controllable devices and/or resources.

As used herein “optimizer input information” includes theinput information600 and the measured/observedinput information604. For purposes of facilitating the description of the invention, theEOS module132 and thesystem ID process602 will be described with reference to a thermal model estimate to be used in the optimization process. However, embodiments of the invention are not limited to a thermal model estimate. TheEOS module132 can generate other models based on the particular application.

FIG. 7 shows a flowchart of the optimization process performed by theEOS module132 according to an embodiment of the invention. Instep700, optimizer input information and a system ID model estimate are received. Instep702, an optimization problem is formulated based on the received input. In the initial phase of energy optimization, the model data is combined with user input and passed to the algorithm as a data structure. This data structure encapsulates an instance of the optimization problem starting from the current time instant up to a point T into the future. The time period T (e.g., a day), is called the prediction horizon.

The problem data structure is parsed by the algorithm and converted into a standard optimization problem (e.g., linear programming, quadratic programming, nonlinear programming, etc.) whose objective is to minimize the cost of energy to the end user. The concatenation of all decision variables in the optimizing problem is denoted by “x”, and contains zone temperature set point commands, grid/PV/battery/HVAC/appliances power flow commands, hot water set point command, and additional (slack) variables for modeling constraints, etc.

The optimization problem, according to embodiments of the invention, includes at least three blocks:

Objective function: f(x), which can contain the following exemplary terms that model:

f1) cost of buying/selling power from/to the grid

f2) cost of operating of heating/cooling systems

f3) cost of operating hot water heater

f4) penalty on violation of user set temperature bounds

f5) penalty on cycling battery state of charge.

Inequality constraints: g(x)≧0, such as:

g1) power limits of the HVAC system

g2) user-set temperature bounds

g3) battery minimum/maximum power

g4) upper/lower limits on battery state of charge

g5) maximum allowable load shedding, etc.

And, equality constraints: h(x)=0, including:

h) power balance between generation and consumption

h) zone heat balance

h) hot water heat balance

h) battery state of charge balance, etc.

The optimization problem is solved instep704. Depending on the formulation of the problem, different solvers can be utilized to compute the vector “x*” that minimizes the objective function f(x) while satisfying the constraints g(x*)≧0 and h(x*)=0. The optimization algorithm re-computes at fixed intervals of time (e.g., 15-30 minutes) in order to receive updates on the state of the customer site (e.g., temperatures in various zones, possible change in predicted loads or weather, etc.). This interval, Δt (e.g., in minutes) is called the sampling interval Δt of the algorithm. In practice, Δt should be chosen small enough that dynamic models in the optimizer can be discretized accurately at this time step.

Instep706, theEOS module132 outputs commands to be used to control utility usage at the customer site so that utility usage is optimized. The raw vector output “x” from the optimization solver is post-processed instep706, and correct values to the output variables are assigned. An exemplary list of outputs from the optimizer to the system is given in Table 1 below.

TABLE 1

		Data
Name	M-Code Name	Type	Size	Units	Description

T	TsetptHVACOut	Float	N_z× N	F	Temperature set
		matrix			point
θ	TsetptDHWOut	Float	N	F	DHW set point
		vector
u_gb	GridToBattOut	Float	N	W	Power flow, grid
		vector			to battery
u_pb	PvtoBattOut	Float	N	W	Power flow, PV
		vector			to battery
u_bl	BattToLoadOut	Float	N	W	Power flow, bat-
		vector			tery to load
q_B	BattSOCOut	Float	N	W-h	State of charge
		vector			of battery
[u_ls1;	LoadShedOut	Float	3 × N	W	Load shed
u_ls2;		matrix			recommended
u_ls3]					in tiers 1-3
u_pg	PVtoGridOut	Float	N	W	Power flow, PV
		vector			to Grid
u_pl	PVtoLoadOut	Float	N	W	Power flow, PV
		vector			to load
u_gl	GridtoLoadOut	Float	N	W	Power flow, grid
		vector			to load
	GridGenOut	Float	N	W	Grid generation
		vector			(= u_gl+ u_gb− u_pg)

Table 1

The exchange of information and command responses will be discussed in more detail below.

FIG. 8 is a flow chart showing the system ID process for generating and updating a system ID model estimate of thecustomer site104. Instep800, the optimization input information and the measured/observedinformation604 are received. Instep802, a customer site thermal model is generated using the system ID process. The system ID process compares the predicted zone temperature from the current site thermal model estimate with the actual zone temperature from the sensor measurements, and updates the thermal model estimate by using the mismatch between predicted and actual zone temperatures. The customer site thermal model is saved instep804, and the system ID model estimate is output for use in optimization instep806.

System identification is the process of building a mathematical model of a dynamic system based on measured data from the system that is being identified. In the context of the exemplary customer site energy management system according to embodiments of the invention, system ID is used within theEOS module132 to estimate a thermal model of each customer site. However, the invention is not limited in this regard, and the system ID can be used to estimate a model based on any set of parameters suitable to the application, as noted above. In this embodiment, the following discrete-time, first order customer site thermal model is used:

T(k+1)=AT(k)+B₁T_a(k)+B₂Q_t(k)+B_3hu_h(k)−B_3cu_c(k)

Where

k is the index of a given time step

T ε Rⁿ^zis the vector of temperatures in each zone (n_z: number of zones)

T_ais the ambient temperature

Q_tε Rⁿ^zis the vector of heat gains (internal, from people or equipment)

u_h(k)ε Rⁿ_zis the HVAC heating load, and u_c(k)ε Rⁿ^zis the HVAC cooling load

A ε Rⁿ^z^×n^z, B₁ε Rⁿ^z, B₂ε Rⁿ^z^×n^z, B_3hε Rⁿ^z^×n^z, B_3cε Rⁿ^z^×n^zare matrices of parameters that need to be identified.

In this embodiment, the system ID process monitors the thermal inputs to the customer site and the resulting variation in the zonal temperatures. The inputs to the system ID model are the measured room temperature (T), measured ambient temperature (T_a), prescribed internal heat gain (Q), measured heating load (u_h), and measured cooling load (u_c). Using this input data the system ID updates the estimated thermal model of the home, the A, B₁, B₂, B_3h, and B_3cmatrices, using the least squares error method, for example. In order for theoptimizer module132 to work effectively, the response of the system ID model must closely predict the thermal response of the customer site. To achieve this goal, a recursive system ID algorithm is used; the thermal model is updated as the input data is received, thereby allowing it to follow changes in the thermal response of the customer site caused by seasonal changes, structural changes, open windows, etc.

Theoptimizer module132 uses the thermal model to calculate the optimum temperature set point for all zones in the customer site so as to minimize the cost of heating or cooling required for maintaining the temperatures within a user defined comfort zone.

According to an embodiment of the invention, the following system ID algorithm is implemented in theEOS module132 to solve for the thermal model estimate. φ is the input matrix to the system ID algorithm; {circumflex over (θ)} is the estimated thermal model parameters comprised of the coefficient matrices of the thermal model; γ is a pre-specified forgetting factor that determines how far back the historical data will be utilized in the system ID algorithm.

{circumflex over (θ)}(k)={circumflex over (θ)}(k−1)+L(k)*[T(k)−{circumflex over (θ)}^T(k−1)*φ(k)]^T

L(k)=1/γ*[P(k−1)*φ(k)]*[1/α+φ^T(k)*P(k−1)/γ*φ(k)]⁻¹

P(k)=1/γ*[I−L(k)*φ^T(k)]*P(k−1)

Where,

{circumflex over (θ)}=[A B₁B₂B_3hB_3c]^T
φ=[T T_aQ_tu_hu_c]^T
γ=forgetting factor
α=1−γ.

The performance of the system ID algorithm is characterized at least by (1) the time it takes the algorithm to converge to the actual model, and (2) the sensitivity of the algorithm to measurement noise. Both of these characteristics are strongly impacted by the choice of forgetting factor. The forgetting factor is between 0 and 1. A small forgetting factor results in an algorithm that quickly forgets the previous parameters. This is good for fast tracking of the model parameters, but causes the algorithm to be highly sensitive to noise. A large forgetting factor results in an algorithm that remembers the previous parameter values, resulting in slow tracking of the parameters, but has the advantage of being less sensitive to noise. The forgetting factor can be set based on the application.

TheEOS module132 also generates a real-time estimate of utility costs as the customer makes adjustments to their preferences on the user interface. This feature can be used to guide the customer by showing (in real time) the cost difference that would result if the customer site were to be run at a different set of preferences.FIG. 9 is a flow chart showing the process for providing a real-time estimate of utility use costs. Instep900, the optimization input information and the measured/observed information604 (including the new user settings) are received. Instep902, the received information is used to run the EOS engine and optimize the future energy use assuming that the new user settings were to be applied. The EOS engine is only run once. According to this embodiment, the output is not used to control the customer site until the user issues a final confirmation to use the new settings (unlike the process inFIG. 7 where the optimization is run periodically and used to update the control signals to the customer site). However, there are other operating processes are possible. Instep904, the new energy use profile is compared with the current profile and the difference in cost/energy use is calculated. The potential cost difference is computed for a fixed time horizon (e.g., one month) into the future. This result is displayed instep906. Thus, the user gets a real-time estimate of the decrease or increase in cost that would result as a consequence of running the customer site differently.FIG. 10 illustrates an exemplary display according to an embodiment of the invention, showingenergy price savings1000.

The smart devices or traditional devices at a customer site can be controlled and coordinated in two interacting groups. In one embodiment of the invention, the energy optimizer of theEOS module132 controls and coordinates the thermostat/HVAC, water heater, and Solar PV and battery storage etc., and all the other devices are controlled and coordinated via the directload control module145. In another embodiment of the invention, all of the devices are controlled by the directload control module145, and in an additional embodiment, theEOS module132 controls and optimizes all of the device operation.

FIG. 11 is a flow chart showing an exemplary flow of device control and coordination between theEOS module132 and theDLC module145. Instep1100, it is determined whether DLC control is enabled for the HVAC or domestic hot water heater (DWH), for example. If yes, then instep1102, it is determined whether the there is a DLC command for the HVAC or DWH. If yes, processing continues to step1104 where the DLC command is executed for the HVAC and/or DWH.Step1106 determines whether the i^thpriority device is a DLC enabled. If yes, then step1108 checks for actual DLC command. If yes, the DLC command is executed instep1110 for the i^thpriority device. If no,step1110 is skipped.Step1112 checks to see if all the DLC commands have been executed for all the DLC enabled devices. If yes, complete the current cycle of the DLC control, otherwise, keep go back tostep1106 for the next DLC device.

If the answer in either of

steps

1100 or1102 is no, then processing proceeds withstep1114 where it is determined whether the EOS Module is enabled. If yes, then the energy optimization process is executed instep1116 to output the optimized thermostat and and/or water heater temperature settings. If the answer instep1114 is no, then processing proceeds to step1106.

The direct load control functionality enhances theEOS module132 by complimenting the energy optimizing function. It includes direct load control analytics, feedback to the utility, load scheduling, usage planning and negotiation. It provides a channel for the utility's demand side management (DSM) to control the loads that are not managed by the optimizer. The user has the ability through the user interface to allow or disallow the direct load control functionality.

Direct load control feedback integrates the utility's direct load control signaling into theEOS module132 and gives the utility control of customer site loads in the event of a utility direct load control signal. The directload control module145 negotiates and plans usage for discrete-operation devices. This function plans and coordinates the operation between devices based on the device constraints, user preferences and priority, and cost of operation, for example.

According to an embodiment of the invention, the appliances or devices at a customer site are ranked in the order of descending energy consumption from the top to the bottom. This is also the default priority ranking for theDLC module145 when a load shedding command is received from the utility and the DLC function is enabled by the customer. The default priority list also takes into account the nature of the devices, the comfort of the customers, and criticality of the devices. Of course, the default priority ranking can be overwritten by the customers per their specific preferences and choices.

A DLC event initiated from the utility is most commonly used for peak load reduction, especially for emergency load shedding. It provides an effective and economic tool for the utility to manage its load/power demand with optimally dispatched or minimized generation resources/spinning reserves. It also enhances reliability of the electric power system to avoid power outages and thus improve customer satisfaction. In addition to the benefits utilities enjoy from DLC, end use customers also benefit from the savings DLC generates. DLC is also relatively simpler to implement and adopt compared with other energy management/optimization algorithms.

The following is a description of exemplary types of information exchanged in the energy management system disclosed herein. Most of the functions and information flow defined and described below apply to smart appliances. For existing traditional appliances, the control application will be predominately discrete on and off controls with a smart outlet plug, for example.

Measurement and monitoring applications provide internal data and status, that include, but are not limited to, solar, wind, and fuel cell states, metering of devices within the site (e.g., customer PHEV), monitoring of local conditions (e.g., temperature, humidity, time, airflow, ambient light level, motion), and monitoring of a device state. These applications provide inputs to theEOS module132 and enable processing, optimization, and action based upon the inputs.

A smart meter provides the following exemplary information to the EOS module132: direct load control request/command, including programmable communicating thermostat (PCT), critical peak pricing (CPP), total electrical energy consummation, consumer alert message, meter ID (e.g., IP address), sub-metering for solar PV and PHEV, utility billing and payment (this can come from the “dashboard and Ethernet interface” as an alternative), power outage/grid status, rebate and other incentives—e.g., peak time rebate (PTR), real-time pricing (RTP), and time of use (TOU). Consumer alert messages prompt users with significant event or pricing changes. They can provide users useful information that can be used to make decisions on reducing energy consumption, increasing efficiency or improving system reliability.

While embodiments of the invention have been described in the context of an electric utility, it will be appreciated by those skilled in the art that the method and system can be used for any utility network.

While only certain features of the invention have been illustrated and described herein, many modifications and changes will occur to those skilled in the art. It is, therefore, to be understood that the appended claims are intended to cover all such modifications and changes as fall within the true spirit of the invention.