Processor-based organic Rankine cycle system for predictively-modeled recovery and conversion of thermal energy

A method for collection and conversion of thermal energy to mechanical energy in a system, comprising:

• receiving system sensor data from a thermal energy recovery system for waste heat recovery comprising operational data associated with environmental conditions and ORC-based operational parameters, the thermal energy recovery system comprising an Organic Rankine Cycle (ORC) comprising a propellant heat exchanger, an expander and a condenser, and one or more heat exchangers comprising a heat medium for transferring waste heat from a combustion device to ORC propellant in the propellant heat exchanger for ORC operation;
• receiving, via a control circuit comprising a predictive module, operational parameters trained on a neural network using node-based pattern classification and regression configured to operate the thermal energy recovery system toward configured target operational values based on operational conditions comprising waste heat recovery and secondary power generation from the ORC, for maximizing net power output in the system,
• processing, via the control circuit, the sensor data to evaluate whether the thermal energy recovery system is projected to deviate from the target operational values within a configured time period;
• and
• transmitting, via the control circuit, control signals to dynamically modify operation of the thermal energy recovery system of the system such that further produced system sensor data meets the target operational values, wherein the control signals are configured to adjust thermal or operational conditions by modifying heat input and heat extraction out of at least two thermal sources comprising thermal fluids, jacket water, exhaust and/or ORC propellant to meet the target operational values.

A system for collection and conversion of thermal energy to mechanical and/or electric energy, the system comprising:

• a thermal energy recovery system for waste heat recovery, comprising an Organic Rankine Cycle (ORC) comprising a propellant heat exchanger, an expander, a condenser, and one or more heat exchangers comprising a heat medium for transferring waste heat from a combustion device to ORC propellant in the propellant heat exchanger for ORC operation;
• sensors, operatively coupled to the thermal energy recovery system and the waste heat recovery system, the sensors being configured to produce system sensor comprising operational data associated with environmental conditions and ORC-based operational parameters;
• a control circuit comprising a predictive module, wherein the predictive module comprises operational parameters for a predictive model trained on a neural network configured to model nonlinear and time-dependent ORC system behavior, the modeling comprising generating vector-based representations of sensor-derived input features and associating such representations with predicted operational outputs corresponding to one or more target operational values, the predictions reflecting waste-heat recovery conditions and secondary power generation from the ORC, for maximizing net power output in the system,
• wherein the control circuit is configured to receive and process the produced sensor data to evaluate whether the sensor data is projected to deviate from the target operational values within a configured time period,
• and wherein the control circuit is configured to transmit control signals to dynamically modify operation of the thermal energy recovery system such that further produced system sensor data meets the target operational values, wherein the control signals are configured to adjust thermal or operational conditions by modifying heat input and heat extraction of at least two thermal sources comprising thermal fluids, jacket water, exhaust and/or ORC propellant to meet the target operational values.