AU2023208068C1

AU2023208068C1 - Machine learning based optimal control system for magnetic continuous variable transmission

Info

Publication number: AU2023208068C1
Application number: AU2023208068A
Authority: AU
Inventors: Thanh Tri Lam
Original assignee: Individual
Current assignee: Individual
Priority date: 2023-03-24
Filing date: 2023-07-24
Publication date: 2024-01-18
Anticipated expiration: 2043-07-24
Also published as: AU2023208068B1

Abstract

A machine learning based optimal control system used to optimally control a system of magnetic variable transmission system for multiple sources of inputs and multiple outputs of mechanical power is developed.

Description

TITLE: MACHINE LEARNING BASED OPTIMAL CONTROL SYSTEM FOR MAGNETIC CONTINUOUS VARIABLE TRANSMISSION FIELD

[0001] It is related to the fields of:

(1) Machine Learning, Optimization, Control Engineering, Mechanical Engineering.

DESCRIPTION

1.1 [0002] Contents of a Machine Learning Based Optimal Control System for Magnetic

Continuous Variable Transmission (MLBOCS for M-CVT)

[0003] The Machine Learning Based Optimal Control System for Magnetic Continuous

Variable Transmission (MLBOCS for M-CVT) is particularly appropriate for harvesting

wind or wave energy as power supplied by waves or winds is variable in a wide range.

[0004] The MLBOCS for M-CVT basically contains the followings:

(1) A Magnetic Continuous Variable Transmission (M-CVT) described below:

(a) A Magnetic Continuous Variable Transmission (M-CVT) has a stationary

member referred to as a stator and three rotors, which are a Mechanical Power

Input Rotor (MPIR), an Electrical Power Input Rotor (EPIR) and a Mechanical

Power Output Rotor (MPOR), to be either coaxial or noncoaxial. Of the three

rotors (the MPIR, the MPOR and the EPIR), there are two rotors called the (first)

pole pair rotor (#301) and the (second) pole pair rotor (#303). Each of the (first)

pole pair rotor and the (second) pole pair rotor has a number of pole pairs. The

number of pole pairs of the (first) pole pair rotor and the number of pole pairs

of the (second) pole pair rotor should be different. The last rotor is called the

Pole Piece Rotor (#302). The Pole Piece Rotor has a number of ferromagnetic

pole pieces. The number of ferromagnetic pole pieces of the Pole Piece Rotor

equals to the total number of pole pairs of both the (first) pole pair rotor and

the (second) pole pair rotor. In addition, gaps between the MPIR, the EPIR and

the MPOR are small enough in order to create the best interactions between

magnetic fields of the MPIR, the EPIR and the MPOR. The MPOR has a MPOR

Magnetic Field created by its pole pairs or pole pieces. The variable rotating

(electro)magnetic field created jointly by the pair of MPIR and EPIR interacts

with the MPOR-Magnetic Field making the MPOR rotated. In addition, the EPIR

also integrates a motor-generator which is able to work either as an electric

motor or a generator depending on mechanical power inputs and outputs. The

motor-generator controllably rotates the EPIR thanks to having controllable magnetic interactions between the stationary member (the stator) and the

EPIR. A type of motor-generators applicable is an induction generator. When

working as a motor, the EPIR uses a source of electrical power which is

converted to mechanical power for rotating the EPIR. Then the mechanical

power of the EPIR combines with the input mechanical power being transmitted

through the MPIR. The combined mechanical power is then outputted by the

MPOR at desired (variable) rotational speeds. The EPIR is controllable and has

a role of changing gear ratios of the M-CVT. Whenever rotational speeds of the

MPOR need to be increased, the EPIR works as an electric motor using extra

electric power supplied. Otherwise, if the input mechanical power of the M-CVT

is excessed, the EPIR works as a generator to convert the excessed mechanical

energy to electricity while it maintains the MPOR rotating at desired (variable)

rotational speeds.

(b) The EPIR can be combined with the MPOR to form an EPIR-MPOR as follows:

the EPIR-MPOR has a motor-generatorand a physical rotor controllably rotated

by the motor-generator which creates controllable magnetic interactions

between the physical rotor and the stationary member (the stator). Thus, the

physical rotor, which is controllable and capable to be used for mechanical

power outputs, is able to work as both an EPIR and an MPOR, referred to as an

EPIR-MPOR.

(c) There is a case that, instead of having three rotors, the M-CVT has only two

(physical) rotors and a stationary member referred to as a stator. Among the

two rotors, a rotor is an MPIR and the other rotor is an EPIR-MPOR. Firstly, the

physical rotor of the EPIR is dismissed and an EPIR-MPOR is formed as described

above. Secondly, magnetic interactions between the three rotors, which are the

(first) pole pair rotor, the pole pieces rotor and the (second) pole pair rotor, are

replaced with controllable magnetic interactions between the two remaining

physical rotors, which form the MPIR and the EPIR-MPOR, are established. It

means that the two remaining physical rotors become a rotor and a stator of a

newly added electric motor, of which its stator is rotated (versus the stationary

member) by inputted mechanical power because the stator is set to be the

MPIR. The newly added electric motor has a (controllable) rotating electromagnetic field, of which its rotational vector (or speed), are regarded as the rotational vector (or speed) of the EPIR which is now to be virtual as it is combined in the EPIR-MPOR.

(d) The M-CVT can be configured with two rotors used for mechanical power inputs

and one rotor used for mechanical power outputs together with a stationary

member referred to as a stator: the M-CVT has two MPIRs (with the two rotors)

and one EPIR-MPOR (with the last rotor). So, the M-CVT is able to combine two

sources of mechanical power inputs to obtain a controllable output source of

mechanical power for rotating a device. Another option is that, the M-CVT can

also be configured with two rotors used for mechanical power outputs and one

rotor used for mechanical power inputs: the M-CVT has two EPIR-MPORs and

one MPIR. Thus, the M-CVT is able to redistribute mechanical power from a

source of inputs to two controllable sources of outputs for rotating two

different devices.

(e) The number of pole pairs are usually putted together in a number of strips

either continuous or noncontinuous. They are strips of pole pairs. Similarly, the

number of pole pieces are also usually putted together in a number of strips

either continuous or noncontinuous. They are strips of pole pieces. The total

number of strips of pole pairs of the two pole pair rotors equals to the number

of strips of pole pieces.

(f) A common way is to arrange all these strips to be parallel together so that the

strips of pole pieces move parallelly between the two paralleled layer of

paralleled strips of pole pairs of the two pole pair rotors. However, this type of

arrangement makes the three rotors rotating non-smoothly as all these strips

are parallel, there are steps while the rotors moving. In order to dissolve this

matter, these strips of pole pieces and strips of pole pairs are arranged

unparallelly: strips of at least a rotor are unparallel to strips of the remaining

two rotors. This solution is also applied for two pole pair rotors interacting

directly without the Pole Piece Rotor included.

(g) Mechanical power outputs of the M-CVT can also be varied and controlled by

controlling rotational speeds of the EPIR(s) with a control system. It is how the

M-CVT changes it gear ratios.

(h) It is notable that the three rotors of the M-CVT can be coaxial (#301, #302, #303) or noncoaxial (#338, #340, #344). If the rotors are coaxial, the M-CVT is called "Coaxial M-CVT". Shapes of the rotors of the Coaxial M-CVT can be either cylinders or cones. If the rotors are noncoaxial, the M-CVT is called "Bevelled M CVT" (Figure (c)). Shapes of the rotors of the Bevelled M-CVT are cones. The Bevelled M-CVT has at least two of the three rotors to be noncoaxial. (2) A Machine Learning Based Optimal Control System (MLBOCS).

1.2 [0005] The Machine Learning Based Optimal Control System (MLBOCS)

[0006] The Machine Learning Based Optimal Control System (MLBOCS) is used to control the M-CVT optimally in order to obtain optimal power or energy output. The MLBOCS integrates a Deep Learning Optimization Algorithm (DLOA) which is used to solve the optimization problem for obtaining optimal solutions. The optimization problem, of which the Objective Function is the total power or energy output of the M CVT, is the Mathematical Model of the M-CVT. The combination of both the optimization problem (the Mathematical Model) and a Deep Learning based algorithm, which is the Deep Learning Optimization Algorithm (DLOA), to solve the optimization problem using Artificial Neural Network (ANN) in order to obtain optimal solutions with regards to time steps, is called the Numerical Model of the M-CVT. The MLBOCS composes of the Numerical Model of the M-CVT and Controllers which require Control Parameters derived from the optimal solutions of the optimization problem (the Mathematical Model). As a result, poweroutputs of the M-CVT are optimally controlled.

[0007] States of the M-CVT are called the M-CVT-States which are expressed by a set of time history M-CVT-State-Data. The M-CVT-State-Data are obtained from a system of M-CVT-State-Sensors which compose the following systems of Sensors and related components of the M-CVT:

(1) The Mechanical Power Input Rotor (MPIR) which has: (a) a system of MPIR-State-Sensors used to collect the MPIR-State-Data. The system of MPIR-State-Sensors has: 1. a system of MPIR-Torque-Sensors used to monitor torques of the MPIR.

II. a system of MPIR-Speed-Sensors used to monitor rotational speeds of the

MPIR.

III. an optional system of MPIR-Accelerators used to monitor rotational

accelerations of the MPIR.

IV. A system of MPIR-Pulse-State-Sensors used to monitor MPIR-Pulse-States

which express states of the MPIR whether it is pulsed.

(b) a set of time history MPIR-State-Data obtained from the MPIR-State-Sensors.

The MPIR-State-Data has:

1. a set of time history MPIR-Torque-Data obtained from the system of MPIR

Torque-Sensors.

II. a set of time history MPIR-Speed-Data obtained from the system of MPIR

Speed-Sensors.

III. an optional set of time history MPIR-Acceleration-Data obtained from the

system of MPIR-Accelerators.

IV. A set of time history MPIR-Pulse-State-Data obtained from the system of

MPIR-Pulse-State-Sensors.

(2) The Mechanical Power Output Rotor (MPOR) which has:

(a) a system of MPOR-State-Sensors used to collect the MPOR-State-Data. The

system of MPOR-State-Sensors has:

1. a system of MPOR-Torque-Sensors used to monitor torques of the MPOR.

II. a system of MPOR-Speed-Sensors used to monitor rotational speeds of the

MPOR.

III. an optional system of MPOR-Accelerators used to monitor rotational

accelerations of the MPOR.

(b) a set of time history MPOR-State-Data obtained from the MPOR-State-Sensors.

The MPOR-State-Data has:

1. a set of time history MPOR-Torque-Data obtained from the system of

MPOR-Torque-Sensors.

II. a set of time history MPOR-Speed-Data obtained from the system of

MPOR-Speed-Sensors.

III. an optional set of time history MPOR-Acceleration-Data obtained from the

system of MPOR-Accelerators.

(3) The Electrical Power Input Rotor (EPIR) which has:

(a) a system of EPIR-State-Sensors used to collect the EPIR-State-Data. The system

of EPIR-State-Sensors has:

1. a system of EPIR-Torque-Sensors used to monitor torques of the EPIR.

II. a system of EPIR-Speed-Sensors used to monitor rotational speeds of the

EPIR.

III. an optional system of EPIR-Accelerators used to monitor rotational

accelerations of the EPIR.

IV. a system of EPIR-Power Consumed-Sensors used to monitor time history

variations of a number of electric currents supplied to a number of coils of

the EPIR for the case that the EPIR works as an electric motor. Properties

of the electric currents need to include phases, voltages, amperages and

frequencies in time history. Time history variations of the electric currents

make the coils creating a variable rotating electromagnetic field which

makes the EPIR (or the electromagnetic field) rotating at variable

rotational speeds. Data collected by the EPIR-Power Consumed -Sensors

are the time history EPIR-Power Consumed -Data.

V. a system of EPIR-Power Generated-Sensors used to monitor electrical

power or energy generated by the EPIR for the case that the EPIR works as

a generator. Data collected by the EPIR-Power Generated-Sensors are the

time history EPIR-Power Generated-Data.

(b) a set of time history EPIR-State-Data obtained from the EPIR-State-Sensors.

The EPIR-State-Data has:

1. a set of EPIR-Torque-Data obtained from the system of EPIR-Torque

Sensors.

II. a set of EPIR-Speed-Data obtained from the system of EPIR-Speed-Sensors.

III. an optional set of EPIR-Acceleration-Data obtained from the system of

EPIR-Accelerators.

IV. a set of time history EPIR-Power Consumed-Data obtained from the system

of EPIR-Power Consumed- Sensors.

V. a set of time history EPIR-Power Generated-Data obtained from the system

of EPIR-Power Generated-Sensors.

(4) A set of M-CVT-Mechanical Power Input-Sensors used to collect time history M-CVT

Mechanical Power Input-Data which express mechanical input power (MIP) of the

M-CVT. It is particular preferred if the mechanical input power is harvested from

waves or winds. In this case, the MIP is a time history function, which is called the

MIP-Function, created by waves or winds. The MIP-Function can be handled either

separately or jointly with the MLBOCS using a machine learning based algorithm

(the Deep Learning Optimization Algorithm (DLOA)) to approximate the MIP

Function. The mechanical power input data of the M-CVT and the MPIR are the

same if the M-CVT has only one input rotor.

(5) A set of M-CVT-Mechanical Power Output-Sensors used to collect time history M

CVT-Mechanical Power Output-Data which express mechanical Output power of the

M-CVT. The Mechanical power outputs of the M-CVT can be variable. The purpose

of controllingthe M-CVT is to obtain time history mechanical outputs (of the M-CVT)

which are closest to a desired time history mechanical power required to supply a

device connected to the M-CVT with the most efficiency. If the M-CVT has only one

output rotor, mechanical power outputs of the M-CVT and the MPIR are the same.

(6) An optional M-CVT-Clutch with M-CVT-Clutch-States expressed by an optional set

of M-CVT-Clutch-State-Data which are obtained from an optional system of M-CVT

Clutch-State-Sensors. The system of M-CVT-Clutch-State-Sensors are used to collect

records of states of the M-CVT-Clutch whether it is engaged or disengaged.

(7) A number of optional M-CVT-Regulating System (M-CVT-RS)s. The M-CVT

Regulating Systems are used to regulate either mechanical power (such as a system

of flywheels) or electrical power (such as a system of Energy Storage). States of the

M-CVT-Regulating System (M-CVT-RS)s are called M-CVT-RS-States. Data of the M

CVT-RS-States, which are M-CVT-RS-State-Data, are collected from the M-CVT-RS

State-Sensors. The M-CVT-RS-State-Sensors optionally contains M-CVT-MPRS-State

Sensors and the M-CVT-EPRS-State-Sensors which belong to the following types of

the M-CVT-Regulating Systems:

(a) Mechanical Power Regulating System (MPRS)s which store and release energy

such as systems of flywheels or any other kinds of systems being used to store

and release mechanical energy. The MPRS applied for the M-CVT is called the

M-CVT-MPRS which has a system of sensors used to collect a set of M-CVT-

MPRS-State-Data expressing the M-CVT-MPRS-States. The set of M-CVT-MPRS

State-Data is obtained from the system of M-CVT-MPRS-State-Sensors. The M

CVT-MPRS-State-Data include M-CVT-MPRS-Control Parameter-Data used to

control operations of the MPRSs.

(b) Electrical Power Regulating System (EPRS)s which store and release energy

generated by the M-CVT such as systems of Energy Storages or any other kinds

of Electrical Energy Storages which are used to store and release electrical

energy. The EPRS applied for the M-CVT is called the M-CVT-EPRS which has a

system of sensors used to collect a set of M-CVT-EPRS-State-Data expressing the

M-CVT-EPRS-States. The set of M-CVT-EPRS-State-Data is obtained from the

system of M-CVT-EPRS-State-Sensors. The M-CVT-EPRS-State-Data include M

CVT-EPRS-Control Parameter-Data used to control operations of the EPRSs.

(8) A system of M-CVT-Power Consumed-Control System-Sensors used to collect time

history M-CVT-Power Consumed-Control System-Data which express electrical

energy consumed for controlling the M-CVT. The system of M-CVT-Power

Consumed-Control System-Sensors contains the system of EPIR-Power Consumed

Sensors and the set of M-CVT-Power Consumed-Control System-Data contains the

set of EPIR-Power Consumed-Data.

(9) A system of M-CVT-Electrical Power Output-Sensors used to collect time history

data of electrical power output of the M-CVT. The data is called the M-CVT-Electrical

Power Output-Data. The electrical power is generated while the EPIR works as a

generator.

(10) An optional system of Temperature-State-Sensors used to collect

Temperature-State-Data.

[0008] The MLBOCS also contains a Master Optimal Control System (MOCS) which is a

Control Software with a Deep Learning Optimization Algorithm (DLOA) integrated. The

MOCS includes the followings:

(1) Input Data, which are in time histories, are collected from different types of

Sensors at a series of time. Some of these data may be derived from the collected

data basing on mechanical or physical relations. The Input Data, which are used to train the DLOA, are described below. Depending on costs of the systems of sensors for data collections and how quick solutions need to be obtained, some of these

Input Data may be optional. The Input Data are:

(a) a set of time history MPIR-State-Data which include a set of MPIR-Torque-Data,

a set of MPIR-Speed-Data, a set of MPIR-Acceleration-Data and a set ofMPIR

Pulse-State-Data.

(b) a set of time history MPOR-State-Data which include a set of MPOR-Torque

Data, a set of MPOR-Speed-Data and a set of MPOR-Acceleration-Data.

(c) a set of time history EPIR-State-Data which include a set of EPIR-Torque-Data,

a set of EPIR-Speed-Data, a set of EPIR-Acceleration-Data, a set of EPIR-Power

Consumed-Data and a set of EPIR-Power Generated-Data.

(d) a set of time history M-CVT-Mechanical Power Input-Data.

(e) a set of time history M-CVT-Mechanical Power Output-Data.

(f) an optional set of time history M-CVT-Clutch-State-Data.

(g) an optional set of time history M-CVT-RS-State-Data which includes an optional

set of time history M-CVT-MPRS-State-Data and an optional set of time history

M-CVT-EPRS-State-Data. The M-CVT-MPRS-State-Data includes M-CVT-MPRS

Control Parameter-Data and the M-CVT-EPRS-State-Data includes M-CVT

EPRS-Control Parameter-Data.

(h) A set of time history M-CVT-Power Consumed-Control System-Data which

include the set of EPIR-Power Consumed-Data.

(i) A set of time history M-CVT-Electrical Power Output-Data.

(j) an optional set of time history Temperature-State-Data.

(k) an optional set of time history Variable-Time Interval-Data. It is optional to

either include or exclude time interval in computation of the Deep Learning

Optimization Algorithm (DLOA). Time interval, which equal to (t(l+1)-t(l)), is a

period of time between two adjacent time steps which are t(l) and t(1+1)

(where i is the index of the time step). If the DLOA optimises energy output,

which reflects total energy generated over a period of time (time interval),

then time intervals are required for the computation. Otherwise, if the DLOA

optimises power output, time intervals are not required for the computation although the time interval does actually exist as the time history Input Data are values at the series of time. Thus, the DLOA can apply either fixed or variable time intervals. The Variable Time Interval is recommended to be used for maximizing energy output and controls. The Variable-Time Interval-Data is obtained from outputs of the DLOA for every time step: the DLOA determines

(predicts) how long the predicted optimal time interval of the Predicted

Optimal Time Step is with the condition that, at the end of the predicted

optimal time interval, the energy output obtained over the predicted optimal

time interval is maximized.

(2) Output data, which are obtained from the optimal solution of the Objective

Function at a Predicted Optimal Time Step (POTS) t(I+1). The Output Data are

required for the Control Policy of the MLBOCS. The Output Data include:

(a) a predicted Optimal Time Interval (POTI), which is (t(I+1)-t()), or alternatively,

the Predicted Optimal Time Step (POTS) t(I+1) at which the Objective Function

is optimal. The predicted optimal time interval is derived from the optimal

solution. The predicted optimal time interval is a control Parameter which

requires the M-CVT shifted to its predicted optimized state of the M-CVT at the

Predicted Optimal Time Step t(I+1). The set of time history Predicted Optimal

Time Steps are called the Predicted Optimal Time Step-Data.

(b) a set of predicted M-CVT-Mechanical Power Input-Data (at the Predicted

Optimal Time Step) expressing the MIP-Function approximated by the Deep

Learning Optimization Algorithm (DLOA). If the M-CVT has only one input,

mechanical power input of the M-CVT and the MPIR are the same.

(c) a set of predicted M-CVT-Mechanical Power Output-Data (at the Predicted

Optimal Time Step). If the M-CVT has only one output, mechanical power

output of the M-CVT and the MPOR are the same.

(d) an optional set of predicted M-CVT-Clutch-State-Data (at the Predicted

Optimal Time Step).

(e) a set of predicted MPIR-Pulse-State-Data (at the Predicted Optimal Time Step)

corresponding to predicted MPIR-Pulse-States which are optionally required

for the Control Policy.

(f) a set of predicted M-CVT-Electrical Power Output-Data (at the Predicted

Optimal Time Step).

(g) a set of predicted Optimal Control Parameter (POCP)s applied at the Predicted

Optimal Time Step: predicted optimal values of variables of the Objective

Function at the Predicted Optimal Time Step t(+1) where the Objective

Function is optimal. These predicted optimal values of variables, from which

the predicted Optimal Control Parameter (POCP)s are derived for:

1. controlling the EPIR: the EPIR has a set of time history EPIR-Power

Consumed-Data and a set of time history EPIR-Power Generated-Data. The

values of predicted EPIR-Power Consumed-Data and predicted EPIR-Power

Generated-Data at the Predicted Optimal Time Step t(+1), which are

obtained from the optimal solution, are the optimal control Parameters

used to control the EPIR at the Predicted Optimal Time Step t(I+1). The EPIR

is controlled to work as an electric motor or a generator.

II. controlling the M-CVT-Clutch: The M-CVT-Clutch, which is optional, has a set

of M-CVT-Clutch-Control Parameter-Data. The values of predicted M-CVT

Clutch-Control Parameter-Data at the Predicted Optimal Time Step t(+1),

which are obtained from the optimal solution, are the optimal control

Parameters used to control the M-CVT-Clutch at the Predicted Optimal Time

Step. In addition, the set of predicted MPIR-Puse-State-Data at the

Predicted Optimal Time Step can also be taken into account for controlling

the M-CVT-Clutch. It is notable that the EPIR can also be used as a Clutch.

III. Controlling the (controllable) M-CVT-Regulating Systems: Control

Parameters for the (controllable) M-CVT-Regulating Systems are obtained

from the predicted Optimal Control Parameter (POCP)s. The values of

predicted M-CVT-MPRS-Control Parameter-Data and predicted M-CVT

EPRS-Control Parameter-Data at the Predicted Optimal Time Step are used

to control the M-CVT-MPRSs and the M-CVT-EPRSs which belong to the M

CVT-Regulating Systems.

(3) The Optional Additional Strategies for the Deep Learning Optimization Algorithm:

The DLOA deploys the following DLOA-Additional Strategies:

Variable Time Interval Method: The optimization problem is solved (or the Objective Function is optimized) at a series of time, which are called time

steps (..., t(I-2), t(I-1), t(I), t(+1), ... ), in which "I" is index, to obtained

optimal solutions. The time steps where the Objective Function is optimal

is called the optimal time steps (..., t(I-2), t(I-1), t(1), t(1+1), ... ). A period of

time between two adjacent time steps is called a time interval. For

example, time interval of the two adjacent time steps t(1) and t(1+1) is Ti(I)

= (t(I+1)-t(I+1)). If the two adjacent time steps are optimal (optimal time

steps), the time interval Ti(l) is called the optimal time interval. If the

control system of the M-CVT has just passed the time step t(l) but has not

reached the time step t(1+1) then t(l) is called the current time step and

t(I+1) is called the predicted time step. In the optimization problem, which

is the Mathematical Model of the M-CVT, time is treated as a variable of the Objective Function. There are two options, which are related to time intervals, to solve the optimization problem. The first option is to treat the time interval (and time step) as Parameters: the time step is predefined and, as a result, the time interval is fixed. In this case, the Objective Function is optimized at a known values of the time step and the time interval. The second option, which is the Variable Time Interval Method, is to treat the time step and, as a result, the time interval to be variable. In this case, values of both the time step and the time interval are derived from the optimal solution of the optimization problem. As a result, values of the time intervals vary with respect to time. The second option of using Variable Time Interval is recommended because if these time intervals are fixed, more time steps may be required for obtaining converged solutions. As a results, more computational tasks and more control operations are executed at every time steps. So, it is optional to optimize the Objective

Function either at a predefined time step (the time t is regarded to be a

Parameter with its value to be the predefined time step of the Objective

Function) or at an optimal time step (the time to is regarded to be a

variable of the Objective Function) which is obtained from the optimal

solution of the Objective Function. The time interval derived from the

optimal time step, which is obtained using Variable Time Interval Method,

is called the optimal time interval. So, the purpose of the optimization

problem is to obtained the optimal solution which is occurred at the

Predicted Optimal Time Step. Data derived from the optimal solution are

predicted data which reflect the next state of the components or portions

of the M-CVT. Control Parameters, which are used to progress (or control)

the M-CVT to the next state, are derived from the predicted data. The

predicted optimal time interval can be obtained as follow: Firstly, the time

(t) is also treated as a variable of the Objective Function. As the Input Data

reflect values at the series of time, the value of these series of time are also

used in training the DLOA with respect to the variable time (t) of the

Objective Function. The optimal solution of the Objective Function

provides the value of the Predicted Optimal Time Step, from which the

predicted optimal time interval is derived. It is also mentioned that the

max/min value of the Objective Function over the period of time [t(),

t(I+1)] equal to max/min (Objective Function (t(l)), Objective Function

(t(I+1)), Objective Function (Optimal Time (t))).

II. Preselected Time Interval Method: It is possible to select a value for time

interval of the next time step t(I+1). As the time interval of the next time

step t(I+1) is known, the Objective Function now can be optimized with

time to to be a constant instead of a variable. In this case, the optimization

process for the Objective Function using Machine Learning (or Deep

Learning) is still the same as presented above.

III. Time Sub-Step: The predicted Optimal Time Interval (of the Predicted

Optimal Time Step) is divided into a number of Sub-Steps. Based on results

obtained at the current optimal time step t(l) and at the Predicted Optimal

Time Step t(1+1), the predicted M-CVT-State-Data of the M-CVT-States at

each sub-step can be approximated. At the meantime, as the time being, data obtained from the system of M-CVT-State-Sensors for each time sub step are collected. Then the predicted M-CVT-State-Data and the collected data of each time sub-step are compared to evaluate accuracies of the progress from the current optimal time step t(I) to the Predicted Optimal Time Step t(I+1). If the accuracies are not good enough, the Predicted Optimal Time Step t(I+1) should be re-established with a new optimization. IV. Interactive Learning Control Method (ILCM): As Data collected from the M-CVT-State-Sensors become more and more, the DLOA can learn newly collected Data and find optimal solutions interactively. V. Untrained Initialized Method (UIM): The Master Optimal Control System of the MLBOCS for M-CVT can be started without prior learning. Firstly, the MLBOCS for M-CVT starts working with a simple control policy which does not need to include optimization. In this case, although the energy generated is not optimized, the control system can be configured to be able to work without optimization while data are being collected and the DLOA is being trained. Once the DLOA is fully trained, it is ready to work fully. In this case, Interactive Learning Control Method (ILCM) should be integrated.

[0009] Construction of the Mathematical Model of the M-CVT

[0010] The Mathematical Model of the M-CVT (the optimization problem) applied to find maximum power or energy output is an optimization problem. Assuming that

mechanical power outputs of the M-CVT is the function MOut(t(l)), of which t(l) is time

step and I is index. If the M-CVT includes Energy Storages used as regulators, the

function of power output of the Energy Storages (SOut(t(l))) is also taken into account.

In addition, power consumed by the Master Optimal Control System (MOCS), including power used for changing gear ratios by rotating the EPIR, is expressed with the Function

EConsumed(t()). Furthermore, power generated by the EPIR in the case it works as a generator is expressed with the function EGenerated(t(I)). Thus, we have the following functions of total power output and total energy output at the time step t():

(1) Total power output of the M-CVT at the time step t(I) is PowerOut(t(I))=

MOut(t(I)) + SOut(t(I)) - EConsumed(t()) + EGenerated(t()), where "I" is the index

of the time step t(I).

(2) Total energy output of the M-CVT over a time interval Ti(), which equals (t(+1)

t(I)), is the function EnergyOut(Ti()) = (Ti()) x (PowerOut(t(I))) where Time

Interval is the period of time between two adjacent time steps t(I) and t(+1). Total

energy output of the M-CVT over multiple Time Intervals is the sum of each total

energy output of the M-CVT over each Time Interval.

(3) Total energy output of the M-CVT over a period of time, which is the sum of

multiple consecutive time interval Ti(l), is the function TotalEnergyOut() where

TotalEnergyOut() equals to the sum of multiple EnergyOut(Ti()).

[0011] Variables of the above functions are time (t(l) and varieties of variables

represented in series of time history Input Data. It is notable that, among variables of

the Objective Function, there are variables which have constraints. For example, the

constraint of gear ratio (equals to ratios of rotational speeds of the MPIR and the

MPOR), which is a type of variables of the Objective Function, is min (gear ratio) < V

Gear-Ratio < max (gear ratio) where V-Gear-Ratio is the variable which represents a gear

ratio.

[0012] Why it is an optimization problem applied to maximize power output?

Depending on devices connected to the output of the M-CVT, requirements for

mechanical outputs may varies. For example, if the M-CVT is connected with a

Generator, the mathematical model of the DLOA is to solve an optimization problem

which maximizes either power or energy output by maximizing the Objective Function

of the M-CVT using Deep Learning Optimizers such as Gradient Descent or Conjugate

Gradient. The Objective Function of the M-CVT is chosen from one of the following

options:

(1) PowerOut(t(I)).

(2) EnergyOut(Ti(I)).

(3) TotalEnergyOut(.

(4) DOutput() minus (PowerOut(t(I)) or EnergyOut(Ti(I)) or TotalEnergyOut() where

DOutput() is a required time history variable mechanical inputs (for best

performance) of a device rotated by the M-CVT.

[0013] Equivalently, the optimization problem can be solved by minimizing the minus

value of the Objective Function. The process of optimization of the Objective Function

can be alternatively done by minimizing either the loss function or the cost function of

the Objective Function. If a device connected to the output of the M-CVT requires a

time history variable mechanical inputs (for its best performance) expressed by a

function called DOutput(, it is possible to minimize the function which equals to

DOutput() minus PowerOut(t(I)) or EnergyOut(Ti(I)) or TotalEnergyOut(.

[0014] Solving the optimization problem (the Mathematical Model of the M-CVT): Why

Machine Learning or Deep Learning are used to solve this optimization problem which

is the mathematical model of complex physical interactions? As described above, the

maximization of the power or energy outputs lead to solving an optimization problem.

In order to solve this optimization problem, Machine Learning or Deep Learning are

applied to adapt complex interactions between devices of the M-CVT while being

subjected by variable mechanical power inputs and temperatures.

[0015] Optimal control of the M-CVT: Why can the M-CVT be controlled optimally?

Optimal control Parameters, which are derived from the optimal solution, are deployed

for the control policy which is used to optimally control the M-CVT. All controllable

devices are started then completed progressing to their new states within allowable

periods of time which are also derived from the optimal solution. It is notable that, if

more accurate optimal solutions are required, the state of the M-CVT for the period of

time while it is being shifted from its current state to the new optimal state is also taken

into account when optimizing the Objective Function.

[0016] How is the Deep Learning Optimization Algorithm (DLOA)? The DLOA is

composed of:

(1) popular deep learning algorithms used to solve the optimization problem (the

Mathematical Model of the M-CVT) to find out optimal solutions with respect to

optimal time steps. The deep learning algorithms includes artificial neural network

(ANN)s with an input layer, multi deep layers and an output layer. As deep learning

algorithms are popular, they are not presented in this document. The deep

learning algorithms are used to find (predicted) approximated functions of the

power or energy outputs of the M-CVT. In other words, the deep learning

algorithm is used to solve the optimization problem which is the Mathematical

Model of the M-CVT.

(2) enhancements of the deep learning algorithm by applying the DLOA-Additional

Strategies.

Claims

1. a Machine Learning Based Optimal Control System for Magnetic Continuous

Variable Transmission (MLBOCS for M-CVT) comprising:

a Magnetic Continuous Variable Transmission (M-CVT); Wherein:

the M-CVT comprising:

o a Mechanical Power Input Rotor (MPIR); where in the MPIR receives

mechanical power inputs;

o a Mechanical Power Output Rotor (MPOR); wherein the MPOR delivers

mechanical power outputs;

o an Electrical Power Input Rotor (EPIR) comprising a motor-generator; wherein:

> the motor-generator is controllably operated by a source of variable

electric power;

> the motor-generator is able to work relying on mechanical power inputs

and mechanical power outputs for its dual functions of driving a device and

generating electricity; wherein:

/ the motor-generator works as an electric motor together with the

MPIR to make the MPOR delivering controllable mechanical power

outputs;

/ the motor-generator works as a generator to generate electricity from

excessed mechanical power inputs while maintaining the MPOR

delivering controllable mechanical power outputs;

o wherein:

> among the MPIR, the MPOR and the EPIR, there are the followings rotors:

a (first) Pole Pair Rotor, a (second) Pole Pair Rotor and a Pole Piece Rotor;

> wherein the pole pair rotors:

v' each Pole Pair Rotor has a number of pole pairs;

v' the number of pole pairs of the (first) Pole Pair Rotor and the number

of pole pairs of the (second) Pole Pair Rotor should be different;

v' the pole pairs are putted together in a number of strips of pole pairs;

wherein each strip is either continuous or noncontinuous;

> and wherein the Pole Piece Rotor:

V the Pole Piece Rotor has a number of ferromagnetic pole pieces;

V the ferromagnetic pole pieces are putted together in a number of

strips of pole pieces; wherein each strip is either continuous or

noncontinuous;

V the number of ferromagnetic pole pieces equals to total pole pairs of

both the (first) Pole Pair Rotor and the (second) Pole Pair Rotor;

/ total strips of pole pairs of the two pole pair rotors equal to the

number of strips of pole pieces;

o and wherein:

> the MPOR has a MPOR-Magnetic Field created by its pole pairs or pole

pieces;

a variable rotating (electro)magnetic field created jointly by the MPIR and

the EPIR interacts with the MPOR-Magnetic Field making the MPOR

rotated;

• and wherein the EPIR of the M-CVT is controlled by a control system to work as

either an electric motor or a generator depending on mechanical power inputs of

the MPIR and mechanical power outputs of the MPOR, including rotational

directions; wherein:

o the control system controls electric currents supplied by a source to rotate the

EPIR for changing gear ratios in order to maintain required mechanical power

outputs of the M-CVT;

o the control system controls electric currents generated by the EPIR while

maintaining required mechanical power outputs of the M-CVT;

a Machine Learning Based Optimal Control System (MLBOCS) comprising:

• a system of M-CVT-State-Sensors;

• a Master Optimal Control System (MOCS);

• wherein the system of M-CVT-State-Sensors comprising:

o a system of MPIR-State-Sensors further comprising: • a system of MPIR-Torque-Sensors used to monitor torques of the MPIR;

• a system of MPIR-Speed-Sensors used to monitor rotational speeds of the

MPIR;

• an optional system of MPIR-Accelerators used to monitor rotational

accelerations of the MPIR;

• a system of MPIR-Pulse-State-Sensors used to monitor MPIR-Pulse-States;

o a system of MPOR-State-Sensors further comprising:

• a system of MPOR-Torque-Sensors used to monitor torques of the MPOR;

• a system of MPOR-Speed-Sensors used to monitor rotational speeds of the

MPOR;

• an optional system of MPOR-Accelerators used to monitor rotational

accelerations of the MPOR;

o a system of EPIR-State-Sensors further comprising:

• a system of EPIR-Torque-Sensors used to monitor torques of the EPIR;

• a system of EPIR-Speed-Sensors used to monitor rotational speeds of the

EPIR;

• an optional system of EPIR-Accelerators used to monitor rotational

accelerations of the EPIR; • a system of EPIR-Power Consumed-Sensors used to monitor time history

variations of a number of electric currents supplied to a number of coils of

the EPIR working as an electric motor; • a system of EPIR-Power Generated-Sensors used to monitor electrical

power or energy generated by the EPIR working as a generator;

o a system of M-CVT-Mechanical Power Input-Sensors used to monitor

mechanical power input of the M-CVT;

o a system of M-CVT-Mechanical Power Output-Sensors used to monitor

mechanical power output of the M-CVT;

o an optional system of M-CVT-Clutch-State-Sensors used to monitor M-CVT

Clutch-States;

o An optional system of M-CVT-RS-State-Sensors comprising:

Sa system of M-CVT-MPRS-State-Sensors used to monitor M-CVT-MPRS

States; a system of M-CVT-EPRS-State-Sensors used to monitor system of M-CVT

EPRS-States;

o a system of M-CVT-Power Consumed-Control System-Sensors used to monitor

electrical energy consumed for controlling the M-CVT; wherein the M-CVT

Power Consumed-Control System-Sensors comprising a system of EPIR-Power

Consumed-Sensors;

o a system of M-CVT-Electrical Power Output-Sensors used to collect time history

data of power output of the M-CVT;

o an optional system of Temperature-State-Sensors;

• and wherein the Master Optimal Control System (MOCS) comprising:

o a set of time history Input Data, known as the M-CVT-State-Data;

o a set of time history Output Data; wherein: • the Output Data comprising a set of time history Predicted Optimal

Control Parameters (POCP);

• the POCP is used to construct the Control Policy of the MOCS;

• the Control Policy is used to control operations of controllable devices of

the M-CVT;

o a Controlling Software; wherein: • the Controlling Software reads the set of Input Data;

• the Controlling Software generates the set of Output Data;

• the Controlling Software instructs the controllable devices of the M-CVT

to progress to new states specified by the set of Predicted Optimal

Control Parameters (POCP); • the Controlling Software comprising the Deep Learning Optimization

Algorithm (DLOA) using the set of Input Data and generating the set of

Output Data;

o a Deep Learning Optimization Algorithm (DLOA); wherein the DLOA

comprising an algorithm of Deep Learning using Artificial Neural Networks to

solve an Optimization Problem; wherein:

• the Deep Learning Optimizers applied can be any kind of popular

Optimizers such as Gradient Descent or Conjugate Gradient;

• the Optimization Problem is a Mathematical Model of the Multiphysics System of the M-CVT; • the Multiphysics System incorporates complex interactions between a number of multi physical portions of the M-CVT such as temperatures, magnetic fields, Regulating Systems and motions of mechanical devices; • the Objective Function of the Optimization Problem of the M-CVT is optionally chosen from the following: > power output of the M-CVT expressed as the function Power

Out(t(I+1)); wherein t(I) is the time step at the time indexed (I) and

t(I+1) is the time step at the time indexed (1+1);

> energy output of the M-CVT expressed as the function Energy

Out(Ti()); wherein Ti(I) is the time interval between the time step

t(I) and the time step t(+1);

> total energy output of the M-CVT over a period of time covering multiple time intervals, expressed as the function Total-Energy-Out(;

> Desired Function() minus Power-Out(t(I+1)) or Energy-Out(Ti(I)) or

Total-Energy-Out; wherein the Desired Function() expresses time history mechanical power inputs required by a device rotated by the MPOR; • the solutions of the Optimization Problem are used to construct the set of Predicted Optimal Control Parameters (POCP); o and wherein the Deep Learning Optimization Algorithm (DLOA) further optionally comprising one or more of the following strategies: • an optional Variable Time Interval Method applied for optimizing the Objective Function with variable time intervals; • an optional Preselected Time Interval Method applied for optimizing the Objective Function with preselected time intervals; • an optional Interactive Learning Control Method (ILCM) applied for learning and optimizing phases of the Deep Learning Optimization Algorithm (DLOA); an optional Untrained Initialized Method (UIM) applied for the initialization of the Master Optimal Control System (MOCS) with an untrained Deep Learning Optimization Algorithm (DLOA); o and wherein the set of time history Input Data comprising:

• a set of time history MPIR-State-Data further comprising:

a set of time history MPIR-Torque-Data obtained from the system of

MPIR-Torque-Sensors;

a set of time history MPIR-Speed-Data obtained from the system of

MPIR-Speed-Sensors;

a set of time history MPIR-Acceleration-Data obtained from the

system of MPIR-Accelerators;

a set of time history MPIR-Pulse-State-Data obtained from the system

of MPIR-Pulse-State-Sensors;

• a set of time history MPOR-State-Data further comprising:

a set of time history MPOR-Torque-Data obtained from the system of

MPOR-Torque-Sensors;

a set of time history MPOR-Speed-Data obtained from the system of

MPOR-Speed-Sensors;

a set of time history MPOR-Acceleration-Data obtained from the

system of MPOR-Accelerators; • a set of time history EPIR-State-Data further comprising:

a set of time history EPIR-Torque-Data obtained from the system of

EPIR-Torque-Sensors;

a set of time history EPIR-Speed-Data obtained from the system of

EPIR-Speed-Sensors;

a set of time history EPIR-Acceleration-Data obtained from the system

of EPIR-Accelerators;

a set of time history EPIR-Power Consumed-Data obtained from the

system of EPIR-Power Consumed-Sensors;

a set of time history EPIR-Power Generated-Data obtained from the

system of EPIR-Power Generated-Sensors;

• a set of time history M-CVT-Mechanical Power Input-Data obtained from the system of M-CVT-Mechanical Power Input-Sensors; • a set of time history M-CVT-Mechanical Power Output-Data obtained from the system of M-CVT-Mechanical Power Output-Sensors; • an optional set of time history M-CVT-Clutch-State-Data obtained from the optional system of M-CVT-Clutch-State-Sensors; • an optional set of time history M-CVT-RS-State-Data further comprising: an optional set of time history M-CVT-MPRS-State-Data further comprising a set of time history M-CVT-MPRS-Control Parameter Data; an optional set of time history M-CVT-EPRS-State-Data further comprising a set of time history M-CVT-EPRS-Control Parameter-Data; • a set of time history M-CVT-Power Consumed-Control System-Data further comprising the set of EPIR-Power Consumed-Data; • a set of time history M-CVT-Electrical Power Output-Data; • an optional set of time history Temperature-State-Data; • an optional set of time history Variable-Time Interval-Data; o and wherein the set of time history Output Data comprising: • a set of time history Predicted Optimal Time Step-Data; • a set of predicted M-CVT-Mechanical Power Input-Data; • a set of predicted M-CVT-Mechanical Power Output-Data; • an optional set of predicted M-CVT-Clutch-State-Data; • a set of predicted MPIR-Pulse-State-Data; • a set of predicted M-CVT-Electrical Power Output-Data; • a set of predicted Optimal Control Parameter (POCP)s applied at the Predicted Optimal Time Step; wherein the POCPs comprising: a set of predicted EPIR-Power Consumed-Data and a set of predicted EPIR-Power Generated-Data containing parameters for controlling the EPIR at the Predicted Optimal Time Step; wherein: / the EPIR is switched between states of working as an electric motor or a generator;

/ as an electric motor, using the predicted EPIR-Power Consumed

Data, the EPIR is variably rotated by handling the controllable

electric currents consumed bythe EPIR in orderto handle required

rotations of the MPOR;

/ as a generator, while mechanical input power (MIP) being

excessed, the EPIR is controlled for both handling required

rotations of the MPOR and generating electricity using the

excessed MIP;

an optional set of time history M-CVT-Clutch-Control Parameter-Data

and an optional set of predicted MPIR-Pulse-State-Data containing

parameters for controlling the optional M-CVT-Clutch at the Predicted

Optimal Time Step;

a set of predicted M-CVT-MPRS-Control Parameter-Data and a set of

predicted M-CVT-EPRS-Control Parameter-Data containing

parameters for controlling the M-CVT-MPRSs and the M-CVT-EPRSs at

the Predicted Optimal Time Step.

2. a Machine Learning Based Optimal Control System for Magnetic Continuous

Variable Transmission (MLBOCS for M-CVT) according to claim 1;

" wherein its Magnetic Continuous Variable Transmission (M-CVT) is further simplified

to have only two physical rotors; wherein:

" the pole piece rotor is excluded while the (first) pole pair rotor and the (second) pole

pair rotor are remained to be included; wherein:

o magnetic interactions between the pole piece rotor, the (first) pole pair rotor and

the (second) pole pair rotor are replaced with controllable magnetic interactions

between the (first) pole pair rotor and the (second) pole pair rotor; wherein the

controllable magnetic interactions are created by a (controllable) rotating

electromagnetic field established between the (first) pole pair rotor and the

(second) pole pair rotor; wherein the rotating electromagnetic field has a

rotational vector, comprising directions and amplitudes of rotational speeds, to be

used as the rotational vector of the EPIR for the control policy; o a Pole Pair Rotor among the (first) pole pair rotor and the (second) pole pair rotor is a MPIR; wherein the MPIR receives mechanical power inputs; o the other Pole Pair Rotor is an EPIR-MPOR; wherein: o the EPIR-MPOR is combined from an EPIR and a MPOR; wherein:

> the EPIR-MPOR has all features of the EPIR and the MPOR;

> the EPIR-MPOR outputs mechanical power in conjunctions with its

controllable rotations assisted by the MPIR and by the motor-generator

comprised in the EPIR;

> the motor-generator of the EPIR-MPOR converts excessed mechanical

power inputs to electricity while maintaining mechanical power outputs of

the EPIR-MPOR;

all remaining features of the MLBOCS for M-CVT remain unchanged.

3. a Machine Learning Based Optimal Control System for Magnetic Continuous

Variable Transmission (MLBOCS for M-CVT) according to claim 1; wherein its Magnetic

Continuous Variable Transmission (M-CVT) is further enhanced to have two MPIRs and an

EPIR-MPOR; wherein:

" the M-CVT comprises a pair of MPIRs and an EPIR-MPOR; wherein:

• each MPIR receives mechanical power inputs separately;

• the EPIR-MPOR is combined from an EPIR and a MPOR; wherein:

o the EPIR-MPOR has all features of the EPIR and the MPOR;

o the EPIR-MPOR outputs mechanical power in conjunctions with its controllable

rotations assisted by the MPIR and by the motor-generator comprised in the

EPIR;

o the motor-generator of the EPIR-MPOR converts excessed mechanical power

inputs to electricity while maintaining mechanical power outputs of the EPIR

MPOR;

• and wherein the M-CVT is used to combine two sources of mechanical power

inputs via the pair of MPIRs to create controllable mechanical power outputs via

the EPIR-MPOR;

" all remaining features of the MLBOCS for M-CVT remain unchanged.

4. a Machine Learning Based Optimal Control System for Magnetic Continuous Variable Transmission (MLBOCS for M-CVT) according to claim 1; wherein its Magnetic Continuous Variable Transmission (M-CVT) is further enhanced to have a MPIR and two EPIR-MPORs; wherein: E the M-CVT comprises a MPIR and a pair of EPIR-MPORs; wherein: • the MPIR receives mechanical power inputs; • each EPIR-MPOR is combined from an EPIR and a MPOR; wherein: o the EPIR-MPOR has all features of the EPIR and the MPOR; o the EPIR-MPOR outputs mechanical power in conjunctions with its controllable rotations assisted by the MPIR and by the motor-generator comprised in the EPIR; o the motor-generator of the EPIR-MPOR converts excessed mechanical power inputs to electricity while maintaining mechanical power outputs of the EPIR MPOR; • the M-CVT is used to distribute mechanical power from a source of inputs via the MPIR to obtain two controllable sources of mechanical power outputs separately via the pair of EPIR-MPORs using the two separate motor-generators comprised in the pair of EPIR-MPORs; E all remaining features of the MLBOCS for M-CVT remain unchanged.

5. a Machine Learning Based Optimal Control System for Magnetic Continuous Variable Transmission (MLBOCS for M-CVT) according to any one of claims from 1 to 4 further comprising a Multiple Contained M-CVT (MCM-CVT); wherein: E the MCM-CVT comprises a number of M-CVTs; wherein: • rotors of each M-CVT are coaxial and the M-CVTs are coaxial; • the most inner M-CVT is contained by an outer M-CVT; • the outer M-CVT becomes the next inner M-CVT; • the next inner M-CVT is contained by the next outer M-CVT; • the process is continued until reaching the most next outer M-CVT; • the most next outer M-CVT is the largest one; • gear ratios of the MCM-CVT equal to multiplications of all gear ratios of the M CVTs;

• each M-CVT of each pair of adjacent M-CVTs shares a common rotor; wherein: o the common rotor is the MPOR of a M-CVT of the pair; o the common rotor is the MPIR of the other M-CVT of the pair; • each M-CVT has systems of sensors and sets of data like the M-CVT comprised in the claims from 1 to 4; E all remaining features of the MLBOCS for M-CVT remain unchanged.

6. a Machine Learning Based Optimal Control System for Magnetic Continuous Variable Transmission (MLBOCS for M-CVT) according to any one of claims from 1 to 5 further comprising a Multiple Bevelled M-CVT (MBM-CVT); wherein: E the MBM-CVT comprises a number of M-CVTs; wherein: • at least one of the M-CVTs has noncoaxial rotors; • each rotor of each M-CVT have magnetic contacts with at least another rotor of the same M-CVT; • each M-CVT has magnetic contacts with at least one of the remaining M-CVTs; • torque capacity of the MBM-CVT might be less in comparison with coaxial M-CVTs; • at least a MPIR rotor of a M-CVT is used for mechanical power input and at least a MPOR of another M-CVT is used for mechanical power output of the MBM-CVT; • each M-CVT has systems of sensors and sets of data like the M-CVT comprised in the claims from 1 to 5; E all remaining features of the MLBOCS for M-CVT remain unchanged.

7. a Machine Learning Based Optimal Control System for Magnetic Continuous Variable Transmission (MLBOCS for M-CVT) according to any one of claims from 1 to 6 with further enhancements for multiple mechanical power inputs and outputs; where in: EA number of MPIRs of a number of M-CVTs are used for mechanical power inputs from a number of different external sources; EA number of MPORs of a number of M-CVTs are used for mechanical power outputs to a number of different external devices; DA number of M-CVTs do not have EPIRs if these have all three rotors used for either mechanical power inputs or outputs; E each M-CVT has systems of sensors and sets of data like the M-CVT comprised in the claims from 1 to 6;

E all remaining features of the MLBOCS for M-CVT remain unchanged.

8. a Machine Learning Based Optimal Control System for Magnetic Continuous

Variable Transmission (MLBOCS for M-CVT) according to any one of claims from 1 to 7

applied for transmitting energy harvested from waves or winds.