Claims:We claim:
1. A method for management of energy consumption based on cloud computing comprising: monitoring of the local devices for energy parameters by collecting the energy consumption of each device and transmitting to a parameter related to the cloud management console; energy management control in a cloud computing platform according to the collected energy consumption of the respective device for the user setting parameters and parameter adjustment mode of the respective scene control of energy consuming devices.

2. The method as claimed in claim 1 wherein the data related to remote devices is analyzed and the optimal energy setting required to be set to the local device is transmitted to the local device from the cloud.

3. The method as claimed in claim 1 wherein the historical data, data from multiple devices and external control factors are analyzed to set the control parameters of the local device.

4. The method as claimed in claim 1 wherein the energy consumption model in the cloud is updated for all the historical data.

5. The method as claimed in claim 1 wherein the transmission of the data to and from the local controllers to the cloud-based management console is via internet or GPRS.
, Description:Cloud Based Learning Platform For Smart Machine Controllers

Field of Invention and Use of Invention:

This invention relates to the field of Proportional Integral-Derivative (PID) Controllers integrated with sensors and intelligent software for cloud based analysis and control. The invention is related the Internet of Things (IoT) field.

The invention is used in effective control and automation of systems for extensive deployment in various forms in many appliances. The invention monitors local parameters, sends the local data over the communication network to a cloud based server which gathers data from other controllers deployed elsewhere, the server analyses the data and provides instructions to the controller based on the learning of the parameters from various controllers.

Prior Art and problem to be solved:

The PID controllers are in general are tiny microcontroller based systems which can take simple decisions based on sensor current, hysteresis and limited storage conditions of the machine itself. The present PID controllers do not take into consideration the learning from other controllers in the same environment but are separated geographically, the external parameters for determining the control value is from a small sub-set of inputs. Thus the controllers are not able to optimize the system to its optimal efficiency.

Problems with prior art:

Stand-alone systems with no statistical analysis inputs from other systems: The prior art suffers from the fact that is no comprehensive system of energy statistics, analysis and management control. The disparate systems work in isolation and do not pass information and analytical outputs to other deployed systems which are of similar nature. Thus, making it impossible to achieve the most optimal allocation of energy.

Stand-alone – no deployment of networked devices: Cloud computing is developed in recent years in network technology. The prior arts suffer from the lack of such technology available earlier. Therefore, the lack of the cloud computing platform renders the prior art with unmanageable network devices, large distributed databases and servers to cater to the new requirements of the networked PID Controllers.

Objects of the invention:

The principal object of this invention is provide a method for management of PID Controllers, on a centralized platform for plurality of sensors and PID Controllers for centralized management control, to achieve maximum energy saving with automatic to achieve better energy efficiency.

Summary of the invention:

The problem to be solved in the present invention is to provide a system for the deployment of sensors/actuators, controllers, central processors and mobile devices connected to each other through a communication system and to provide configuration, monitoring and control of physical parameters and external conditions.

The invention is Monitoring/logging of information from all the appliances which helps tracking and centralized auditing.

In the invention reverse information fed into the cloud based device controller which can take decisions based on external parameters from other sensors and probes which are not just in the current appliance. In such a scenario there are two kinds of parameters that the local controllers benefit from

Machine learnt appliance parameters based on history: Based on a region or a nature of similar appliances many factors can be back fed into the controllers in the local machine to provide optimizations. For example, a typical decision could be based on the power consumption of the local appliance, which even though could be hitting a high consumption could be attributed to a bad sensor reading which is making the local controller unable to initiate defrost cycles in a refrigeration unit. However the cloud could in such cases compare this power with historic power and initiate the controller to start a defrost cycle which could solve the problem of power consumption and also potentially frosted sensor which now start giving correct value. The cloud could also observe the correction and see if the behaviour is moving in the expected direction of the decision and reverse it in case unnecessary or incorrect and alert users for the same.

Appliance parameters across multiple appliances to take a decision: The local appliance decision may also be driven by global appliance data. Typical cut off cycles of compressors are localized decisions, however based on power data and data of coolant circulation cycles the cloud could help learning the ideal/optimal data correlations for minimal power and drive that algorithmic decisions back to the local controllers for the decisions.

Appliance decisions based on External Control Factors: Tying to sensors across various systems to make global decisions. The cloud could aggregate information of different sensors and using heuristics drives decisions on the appliance locally. For example: we enable a wide range of sensors that can convey information from different sources given to the cloud, a door sensor indicating the store door being open could help the local PID controller to know that the poor efficiency of an air conditioner is expected and take a decision on compressor cut-off cycles accordingly.

Brief Description of Drawings:

Fig 1 illustrates the typical PID Controller system;

Fig 2 is transforms the PID to a transmission device for monitoring, analysing and taking global decision at a cloud level based on deep learning.

Detailed Description:

The detailed description of the invention will be described in detail below with reference to the drawings. FIGS. 1 and 2, discussed below, and the various embodiments used to describe the principles of the present invention in this patent document are by way of illustration only and should not be construed in any way to limit the scope of the invention. Those skilled in the art will understand that the principles of the invention may be implemented in any type of suitably arranged system.

Fig 1. Illustrates the typical PID Controller System.

The connected device which taps into the internal monitoring parameters of any appliance, monitors, analyses and sends alerts to a cloud based on the parameters obtained from the cloud.

The device has digital or Analog ports to monitor and control the parameters of appliance. The device has a capability to make local decisions to control the parameters. The device connects on reliable GPRS or any communication network to perform the below operations
a) Synch with the cloud
b) Upload data to the cloud on the parameters that are being monitored at an interval specified by the cloud
c) Alert in case of the parameter overshoot based on cloud inputs
d) Configure the parameters and IOs based on cloud
e) Take control decisions based on the cloud inputs
f) Ability to upload data reliably in an extremely fine grain requirement of parameters

Fig 2 illustrates the cloud based invention.

The cloud based analytics engine learns from global data parameters and helps improving the local decisions related to alerting and predicting failures in the devices.
Specific requirements from the Cloud
a) Global data base of devices and parameters.
b) Ability to manage/store/analyse and learn the device parameters in an extremely low latency manner.
c) Ability to provide and correct anomalies or predictive decisions to correct any system based on global information.
d) Ability to analyse structure databases for optimal latencies.

The cloud based analytics engine operates on an extremely efficient data store. Data sorted and stored in a highly efficient manner to reduce the latency of fetch and periodically summarized for easier analysis and lighter weight analysis.
b) A high speed ML engine which is looking at finding the patterns on the curve and upon each update from any child appliance looks for rules which may be applied and fed back to the system based on the same device’s data and global parameters.
c) A neural network engine constantly learning the default optimal parameters based on the rules given to it. The key parameters that are tuned are:
a. Sensor details data
b. Optimizing parameters
c. Outcome expectations
Based on the above the device performs and builds a neural network of learning which start optimizing the devices for the expected outcomes globally and feed back inputs to individual devices upon the receipt of their cloud updates or synchs.

The invention has works continuously in tandem/synch with the cloud and updates sensor data to it with minimal analysis. It Receives decisions from the cloud which are based on global parameters from appliances of similar kind, sensors of different kinds and external parameters. The system is able to bulk upload of a large amount of data in chunks of larger intervals in order to operate under un-reliable conditions. A large amount of continuous server transactions are avoided by this feature and also the only way devices can operating without a heavy load on the cloud servers. It has a highly efficient mechanism for alerting, updating granularity of data sends and controlling devices at a fine granularity. The centralized parameter setting for appliances which includes, software based calibration of optimum region or requirement based settings, dynamically changing settings using advanced machine learnt algorithms.