WO2009111478A2 - Machine battery packs and controls - Google Patents

Abstract

One aspect of the invention provides devices that comprise a battery pack with switched modules and methods that utilize the same. The electrical connectivity of the modules and the battery cells that comprise the modules can be changed based on the desired battery pack characteristics. Modules can be connected or disconnected to a power bus using electrically controlled switches to provide for a range of power supply to a load. The invention also provides for using the battery pack with switched modules to supply power to an electric motor. Another aspect of the invention provides systems and methods for control of switched reluctance machines, which may include formulating a model for the switched reluctance machine that can be optimized using convex optimization algorithms. Control of the switched reluctance machine may be optimized for a desired optimizing objective. The solution to the convex optimization problem may enable a controller to control a power converter for given motor velocities and torque demands, and thereby control a motor.

Description

MACHINE BATTERY PACKS AND CONTROLS

CROSS-REFERENCE

[0001] This application claims the benefit of U.S. Provisional Application No. 61/033,657, filed March 4, 2008 and U.S. Provisional Application No. 61/044,766 filed April 14, 2008, which applications are incorporated herein by reference.

TECHNICAL FIELD

[0002] The invention is related to battery systems and to electrical connections for variable power supply. [0003] Furthermore, this invention relates to systems and methods for controlling a switched reluctance machine using a model that can be optimized using convex optimization algorithms. The solution to the convex optimization problem may be stored as look-up tables determining power converter switch positions for given motor characteristics such as velocity or torque demand.

BACKGROUND OF THE INVENTION [0004] Battery Packs

[0005] Typically, rechargeable battery packs utilize a group of battery cells as a single unit without control over connectivity of individual batteries. These types of battery packs can have a number of deficiencies. In particular, voltage supply is difficult to control, repair of the battery pack can be challenging, cell failure is catastrophic to battery pack operation, and charging of the battery pack can be performed without regard to the characteristics of each individual battery cell.

[0006] The voltage output of a typical battery pack is dependent on the amount of charge remaining. The voltage output is highest when the battery is fully charged and gradually drops as the battery charge is depleted. In this case, the user cannot control the battery pack voltage for efficient usage of the battery pack without using a separate voltage regulator. [0007] Another characteristic of battery cells is that they tend to degrade unpredictably and can cause battery packs to suffer from undesirable performance characteristics. When individual battery cells of battery packs are faulty, the battery pack is often unusable and must be disassembled in order to be repaired. This requires the battery pack to be taken offline and interrupts power supply to the load. In addition, the disassembly and reassembly process is time consuming and difficult to perform.

[0008] The variation between individual battery cells within a battery pack can also cause undesirable charging characteristics. In some cases, individual battery cells in a battery pack may have differing charge capacity and some battery cells may reach full capacity during charging before others. Continued charging of a battery that has reached full capacity is undesirable and can damage the performance of the battery pack.

[0009] Because of these reasons, there is a need to design a battery pack that can be capable of controlling of the electrical connectivity of individual battery cells, allowing for control of battery pack voltage, on-line battery repair, and more desirable charging conditions. [0010] Control of Switched Reluctance Machine

[0011] A switched reluctance machine (SRM) may include a switched reluctance motor with a rotor without conductive windings or permanent magnets and a stator with electronically- switched windings carrying unidirectional currents. Commonly, pairs of diametrically opposed stator poles may be connected in series or parallel to form one phase of a potentially multi-phase switched reluctance motor. Motoring torque may be developed by applying voltage to each of the phase windings in a predetermined sequence that may be synchronized with the angular position of the rotor so that a magnetic force of attraction results between poles of the rotor and stator as they approach each other. Depending on the number of rotor and stator poles there may be a number of phase periods for each complete rotation of the rotor. [0012] The state of a SRM is the minimal collection of all information necessary to entirely characterize the SRM at that moment. For example, the SRM 's behavior may be deterministically predicted if the rotational speed, rotor position, flux linkages and phase currents are known. These four quantities characterize the state of the SRM. Alternatively, the instantaneous state (on or off) and a finite history of the states of each of the switches of a power converter that provides the power supply to the motor can be used to characterize a SRM instead of flux linkage and phase currents. Some power converters may have two switches per phase, while others may have one or more switches per phase.

[0013] Various strategies have been proposed in the past for controlling switched reluctance motors as part of an overall variable speed drive system. Currently, techniques like open- loop sensorless control, closed-loop torque control, and observer-based control are used to control SRMs. Sometimes closed- loop torque control is embedded within closed- loop velocity control. With the exception of observer-based control and some sensorless control techniques, most control methods use proportional-integral-derivative (PID) control techniques on the feedback loops. Observer-based control methods may model the system in a series of state variables and use the model in a real-time feedback loop. [0014] The relationship between a required value of a motor's torque and the appropriate switch positions for various angular rotor positions for each speed of the motor (assuming a constant DC link voltage) can be stored in a look-up table (LUT). In some known controllers for switched reluctance motors, the LUT may comprise a circuit that can include turn-on and turn- off angle information for various rotor speed and torque demand combinations. In most systems the information that is stored in the LUT is derived empirically. The empirically derived information may be stored in the LUT, sometimes together with information from non-tested speeds and torque demands which has been interpolated from the empirically derived information.

[0015] A need exists for systems and methods of controlling a SRM for an optimized characteristic using a model-based application that can be solved quickly. Using a model-based approach may allow exhaustive searching of the design space that would be infeasible using empirical techniques. Empirical control strategies have no way to verify their optimality and it is not possible to test every single imaginable control strategy empirically. For example, most empirical control strategies assume a single turn-on and single turn-off for each switch over the course of one rotor revolution. Without this assumption, empirical control strategy would not know where to begin looking for all optimal switching times.

[0016] Creating a model and using numerical convex optimization techniques, may enable certainty that switching times are optimal out of all possible switching times. Some empirical information may be used to build a model, but then the model may be solved over a wide range of operating speeds, temperatures, loads, etc. Since SRMs have a highly nonlinear response to current, the ability to solve the system optimally at a wide range of operating characteristics is preferable to using a few empirical guesses and then interpolating. Using a model may enable finding non-obvious control strategies that would never have been tried with empirical guessing. Such a control strategy may drastically reduce torque ripple, increase efficiency, or increase the amount of torque that can be delivered. SUMMARY OF THE INVENTION

[0017] Battery Packs

[0018] One aspect of the invention provides devices that comprise a battery pack with switched modules and methods that utilize the same. Various aspects of the invention described herein may be applied to any of the particular applications set forth below or for any other types of power supply systems or methods. The invention may be applied as a standalone system or method, or as part of an integrated arrangement for powering electric motors, or as part of an integrated arrangement for powering switched reluctance machines. It shall be understood that different aspects of the invention can be appreciated individually, collectively, or in combination with each other. [0019] The battery pack with switched modules may comprise a power bus for carrying current from a variable power source, where the variable power source can include at least one module with a switch that may electrically connect the module to the power bus on a selective basis, and further where a module may include one or more battery cells.

[0020] In one embodiment of the invention, the battery pack may comprise at least two modules that can be arranged in parallel, in series, or some combination thereof. The battery pack may comprise modules of a single type or multiple types. Module type can be determined by the type of battery cell used and by the electrical connections between the battery cells of the module.

[0021] A battery cell used in a module can have a variety of different battery chemistries.

These battery chemistries can include lead-acid, nickel, or lithium based chemistries. A module can be comprised of battery cells of one or more types of battery chemistry.

[0022] In other embodiments of the invention, the battery cells within a module can be in a combination of parallel and serial arrangements, a parallel only arrangement, or a serial only arrangement. The arrangement of battery cells can change the power output and charging characteristics of a module. [0023] In an alternate embodiment of the invention, the battery pack may comprise one module that can be used for pulsing voltage to the power bus, or can be used if other modules have failed or have been removed.

[0024] The power bus voltage can be controlled by changing the state of one or more module switches, wherein the state of the switches can be set by a controller. A module can be intermittently connected to the power bus, while the power bus is able to supply a continuous or constant level of power by intermittently connecting another module. In one embodiment of the invention, a module can be removed without disrupting power supply. The state of a module can be monitored to help determine how to produce a desired power supply. The state of the module can be monitored using devices known to those skilled in the arts. [0025] In another embodiment of the invention, a module can be selectively charged by an external power supply or by another module. The charging of the module can be performed via the power bus.

[0026] The method of using the battery pack with switched modules may comprise controlling the power output of the battery pack by setting the state of one or more switches that can electrically connect one or more modules to a power bus. In other embodiments of the invention, the method of using a battery pack with a switched module may comprise controlling the charging of a module by setting the state of one or more switches that electrically connect the module to a power bus. Monitoring the state of the module can help determine a discharging or charging procedure.

[0027] In another embodiment of the invention, the battery pack with switched modules can be used to power electric motors. The battery pack with switched modules can supply a range of voltage and current to an electric motor such that the electric motor can operate at a wide range of speeds and produce a wide range of torque. [0028] In some embodiments of the invention, the battery pack with switched modules can be used to power an electric motor that can be used to mechanically drive a wheel of a vehicle. An electric motor used to drive a wheel of a vehicle can be required to operate at a wide range of speeds and produce a wide range of torque. In one embodiment of the invention, the battery pack with switched modules can supply variable power to an electric motor that can be used to mechanically drive wheels of a vehicle, wherein the electric motor can operate at a wide range of speeds and produce a wide range of torque. [0029] Control of Switched Reluctance Machine

[0030] In accordance with another aspect of the invention, a system of a switched reluctance machine (SRM) may be modeled as a series of state variables within a state space. Measurements of various physical properties of a motor may be taken for a motor design. The motor may be a switched reluctance motor, or another type of brushless motor. Such measurements may include the inductance of each phase and mutual inductance between phases for various rotor positions of the motor, and may yield inductance values for each phase at each measured rotor position. The measurements may also be carried out for a motor design over a range of environmental conditions, such as temperature. These measurements allow for generation of operating rules that may vary over temperatures ranges.

[0031] The model, formulated with the empirical measurements of various physical properties of an SRM, may be formulated as a numerical convex optimization problem for a given demand (such as torque demand) and motor operating characteristic (such as velocity). Optimality may be defined when formulating the optimization problem for a desired optimization objective. For example, a desired optimization objective may involve achieving the best possible trade-off between efficiency, torque ripple, and other features of merit. In other examples, a desired optimization objective may involve maximizing efficiency, minimizing torque ripple, minimizing torque delivery time, or optimizing other characteristics. The numerical convex optimization problem may be solved using readily-available numerical techniques. An advantage of formulating the problem as a numerical convex optimization problem may be that numerical techniques for solving the problem can be relatively quick and simple. Also creating a model and using numerical convex optimization techniques may enable certainty that switching times are optimal out of all possible switching times. Using a convex optimization model may enable finding non-obvious control strategies that would never have been tried with empirical guessing.

[0032] A SRM may have a power converter, which may include switches (such as a multiphase asymmetric bridge converter which may consist of two switches and two diodes per phase). The switch positions may be changed to apply power to the phases of the motor. Solving the numerical convex optimization problem may determine the state of each switch of the converter for a given angular position of the motor's rotor to optimize for a desired optimization objective for a given demand and motor operating point. [0033] The switch states may be saved in the SRM controller as a look-up table (LUT) for a given demand and motor operating characteristic, which may include motor torque demand and motor velocity. For example, velocities may range from zero up to the vehicle's top speed. Torques may range from zero up to the vehicle's maximum torque. LUTs for switch states can also be calculated for situations with negative torque (such as for regenerative braking), and for situations with negative velocities (such as operating in reverse).

[0034] Each LUT may tell the SRM how to run for each velocity and each torque demand. In an LUT, the controller would find an array of numbers to serve as an operating rule. The operating rule may tell the controller when to switch on and off all of the converter's switches based on the motor's rotor position, to achieve the desired torque at the given velocity for a given optimization objective (such an objective may be to run as efficiently as possible while ensuring torque ripple remains below a threshold). Each LUT may also have multiple operating rules to be used for different environmental conditions. For example, one environmental condition may be the temperature of the motor. If the SRM temperature were to change, and its inductance were changed, a different rule can be used. In this case, temperature sensors could be attached to the SRM to constantly evaluate which part of the LUT it should be using to operate. Environmental conditions may also include other factors such as ambient vibration level or motor age. [0035] There may be symmetry within the multiple modes of operation, so a new LUT may not need to be generated for every situation. Once the LUTs are stored in the controller, the controller can use them, plus empirical information such as velocity from a motor sensor and torque demand from user input, to determine when to commute SRM phases. [0036] In an alternate embodiment of the invention, the model may be formulated and solved as a convex optimization problem in real-time such that an embodiment would not require a LUT. Instead, a power converter may be controlled by the controller which may formulate and solve the model based on sensor readings and system demands. [0037] Other goals and advantages of the invention will be further appreciated and understood when considered in conjunction with the following description and accompanying drawings. While the following description may contain specific details describing particular embodiments of the invention, this should not be construed as limitations to the scope of the invention, but rather as an exemplification of preferable embodiments. For each aspect of the invention, many variations are possible as suggested herein that are known to those of ordinary skill in the art. A variety of changes and modifications can be made within the scope of the invention without departing from the spirit thereof.

INCORPORATION BY REFERENCE [0038] All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference.

BRIEF DESCRIPTION OF THE FIGURES [0039] The novel features of the invention are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present invention will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the invention are utilized, and the accompanying drawings of which: [0040] Figure 1 is a schematic of a battery pack with switched modules.

[0041] Figure 2 is a schematic showing electrical circuits for variable bus voltage.

[0042] Figure 3 is a schematic showing electrical connectivity for independent discharging and charging of a battery module.

[0043] Figure 4 is a schematic showing a variable power source with an optional capacitor. [0044] Figure 5 is a schematic showing individual cells in a battery module.

[0045] Figure 6 is a schematic showing electrical connectivity of a battery module to a power bus, using an inductor to limit current flow between module and bus. [0046] Figure 7 shows a switched reluctance machine and how a controller, power converter, and switched reluctance motor may interact.

[0047] Figure 8 illustrates information flow between a controller, a power converter, and a motor. [0048] Figure 9 illustrates a method of controlling a switched reluctance machine.

[0049] Figure 10 illustrates a method of controlling a switched reluctance machine for a desired optimization objective.

[0050] Figure 11 illustrates an alternate method of controlling a switched reluctance machine.

DETAILED DESCRIPTION OF THE INVENTION

[0051] While preferred embodiments of the present invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. Numerous variations, changes, and substitutions will now occur to those skilled in the art without departing from the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in practicing the invention.

[0052] I. Battery Packs [0053] A. Overview [0054] One aspect of the invention provides for a supply of power using a battery pack with switched modules and methods of use thereof. As shown in Figure 1, the battery pack may comprise a power bus and a variable power source, wherein the variable power source may include at least one module with one or more switches that can electrically connect the module to the power bus. A module may be comprised of one or more battery cells. [0055] The battery pack with switched modules can comprise a means to monitor the status of the modules. The modules can be monitored for a variety of parameters. For example, the modules can be monitored for voltage, capacity, temperature, or power. A range of devices can be used to monitor the status of the module. For example, these devices can include devices to measure voltage, current or temperature. Devices that could be used to measure voltage include a voltmeter or a two-resistor voltage divider and A/D converter. Devices that could be used to measure current include an ammeter or a Hall Effect sensor. Devices that could be used to measure temperature include a thermocouple or a thermistor. The device used to measure temperature can be a thermistor bundled with an amplifier and an A/D converter in a small IC package. The devices required to monitor the status of the module can be any known to those skilled in the art for monitoring the status of a power supply. [0056] In one embodiment of the invention, the battery pack may further comprise a controller that can change the state of one or more switches that can control the electrical connectivity of one or more modules. The power source can be varied by controlling the state of a switch that can electrically connect a module to the power bus. Control of a switch can also allow for selective discharging or charging of a module. One example of a switch is a buck converter, a boost converter, or a buck/boost converter. This switch may allow for selective charging or discharging of a module. Such a switch may use passive electrical components such as capacitors and inductors to allow better control of power moving into or out of a module. Because the module can be connected and disconnected independently, the module can be removed without reducing the performance characteristics of the battery pack. [0057] Another aspect of the invention provides for methods to use a battery pack with switched modules. In some embodiments of the invention, the method to use a battery pack with switched modules may include receiving information regarding power demand, measuring the status of one or more modules, and using a controller to set the state of one or more switches that can control the electrical connectivity of one or more modules. The method to use a battery pack with switched modules may include monitoring the status of the modules to help determine how to charge the modules. In other embodiments of the invention, the controller may set the state of the switches based on a desired power bus voltage or current. In one embodiment, passive components may be added to the switching circuit to better regulate power bus voltage or current flowing between module and power bus.

[0058] B. Electrical Connections and Control of the Same

[0059] As shown in Figure 2, a switch can be used to control the electrical connectivity of a module to a power bus and can allow for a module to be connected or disconnected from the power bus. In one embodiment of the invention, a switched module can be intentionally or unintentionally removed from the battery pack without interrupting power supply. The immediate performance of the battery pack may not be degraded upon such a removal. [0060] In another embodiment of the invention, an individual module can be intermittently connected a power bus, while the power bus may provide a continuous power supply to a load by intermittent connection of another module, thereby not disrupting power supply to the load. The power bus may also supply a constant level of power to a load by intermittently connecting multiple modules. Passive electrical components may be used to smooth the power bus voltage or current flowing to or from modules. Many battery chemistries, including Li-ion, may respond favorably to intermittent loads, which may increase module power, capacity, and/or lifetime. Additionally, by unequal distribution of load to a module, the module may be discharged more than other modules connected to the power bus. Unequal levels of discharge among the modules may be useful if some modules are higher capacity or higher power. Unequal levels of discharge can be possible because a module can be electrically isolated from other modules. [0061] In some embodiments of the invention, a switch can control the electrical connectivity of a module to control the power bus voltage. The electrical connectivity of the module can then be tailored to provide a desired power supply by using information regarding the state of the module and by using equations known to those skilled in the arts. In one embodiment, an inductor may be used in the switching circuit to better control current flow to or from the module. Capacitors may be used to better control the voltage at each module's output. Equations to control current flow through an inductor or voltage across an inductor are known to those skilled in the art.

[0062] Alternatively, the power bus voltage can be altered by pulse width modulation (PWM) of one or more modules. The pulse width modulation duty cycle can have various shapes, including square, saw-tooth, triangular, sinusoidal, or some combination thereof. PWM may be used on the switches to regulate power flows to or from a module. Passive electrical components may assist in PWM regulation of power flows. The output can then be calculated using information regarding the state of the module and by using equations known to those skilled in the arts. [0063] In other embodiments of the invention, the control of electrical connectivity of a module to a power bus can allow for a wide range of voltage and current to be supplied to a load and may allow for more efficient usage of the battery pack charge. In one embodiment, a buck converter may be used as a switch circuit to allow a lower voltage to be delivered to the load. In another embodiment, a boost converter may be used as a switch circuit to allow a higher voltage to be delivered to the load.

[0064] The control of a power bus voltage by selective connectivity of a module to the power bus can offer a number of advantages over only using pulse width modulation or another voltage regulator to control a supply of power to a load. These advantages can include the ability to supply a greater range of voltage and current to a load, the ability to more appropriately match slowly- varying changes in power demand from a load and the ability to reduce harmonic effects between the supply of power and the load.

[0065] Several different types of electrical components can be used as a switch, such as a transistor, an on-off switch or a relay. The switch can be selected from the group consisting of an insulated-gate bipolar transistor, a field-effect transistor, a junction gate field-effect transistor, a metal oxide semiconductor field-effect transistor, or a bipolar junction transistor. The switches can be purchased from suppliers or custom fabricated. Multiple switches may be used to control power flows between a module of battery cells and the power bus. An example would be a bidirectional buck/boost converter shown in Figure 6.

[0066] As shown in Figure 2, the electrical connections can include a diode to control the directionality of current through a switched electrical connection for discharge of a module. Because modules can be used independently and modules can have variable performance characteristics, a module can be at a lower power charge, as compared to other modules connected to the power bus. In this case, a diode can prevent inadvertent charging of a module that has reduced charge, as compared to other modules connected to the power bus. [0067] In other embodiments of the invention, the switch and diode for controlling discharge of a module can be replaced with a reverse-blocking transistor. The reverse-blocking transistor can be a reverse-blocking insulated-gate bipolar transistor.

[0068] As shown in Figure 3, a switch can be used to selectively charge a module. The same power bus that can be used to discharge a module can be used to charge a module. The electrical connectivity of the battery pack with switched modules can be designed such that a module can be either discharged at one instance or charged at one instance. Selective charging of a module can allow for a depleted battery to be charged at a desired instance, independent of the charging or discharging of other modules. Selective charging of a module can avoid detrimental effects caused by overcharging or improper charging. Because a module can have a different charge capacity as compared to other modules connected to the power bus, a module may reach full capacity before others. In these cases, continued charging of the fully charged module could be detrimental to the performance of the module.

[0069] The charging procedure for a module can be performed by using an external power source, another module, or some combination thereof. In some embodiments of the invention, the charging procedure can be performed by continuous or intermittent electrical connectivity of the module to a power source via one or more switches and passive electrical components. The charging procedure can be performed by electrically connecting the module to a power source via the power bus. In one embodiment of the invention, the charging procedure can be performed using pulse width modulation.

[0070] In other embodiments of the invention, the electrical connections for charging a module can include a diode to control the directionality of current through the switched electrical connection. The diode avoids inadvertent discharging of a module during a charging procedure. [0071] In one embodiment of the invention, the switch and diode for charging can be replaced with a reverse-blocking transistor. The reverse-blocking transistor can be a reverse- blocking insulated-gate bipolar transistor. [0072] In an alternate embodiment of the invention, the switch for discharging and the switch for charging can be combined in a single switch. An example of a single switch that can control current flow to one of two routes is a two-way switch or a single pole, double throw switch. [0073] C. Power Supply

[0074] The power supply can comprise one or more modules connected to a power bus. Shown in Figure 4, the power supply can comprise one or more capacitors. In other embodiments of the invention, the battery pack with switched modules can be electrically connected to a power generator. The elements of the power supply, including the modules, the capacitor or the power generator, can be in a parallel only arrangement, a serial only arrangement, or some combination thereof. [0075] The battery pack with switched modules can be comprised of more than one module, wherein modules of different types can be used in a single battery pack. A module can be differentiated from other modules based on the specific capabilities of the module. For example, a module can be designed to have maximal power capacity, discharge rate, lifetime, energy density, or durability. In some embodiments of the invention, a battery pack can be comprised of any of the following types of modules selected from the group consisting of a high-energy module, a high-power module, a high-voltage module, a low-voltage module, a high capacity module, a low-capacity module, a Li-ion module, and a Li-Fe-P module. A number of factors can determine the characteristics of a module. For example, the characteristics of a module can be dependent on the battery cells that comprise the module and the electrical connections between the battery cells.

[0076] A variety of battery types can be used in a module. The type of battery cell used in the module can impact the capabilities of the module. The different types of battery cells can be differentiated based on a number of factors. These factors, for example, can be voltage, energy density, power capacity, efficiency, cost, discharge rate, charge rate, power density, chemistry, shape, durability, or lifetime. [0077] Battery chemistries of different types can be used in the battery cells. For example, the types of battery chemistries that can be used include lead-acid, nickel, or lithium. A lead- acid battery cell can include a standard lead-acid battery with lead and lead oxide electrodes, a valve regulated lead-acid battery that can have reduced maintenance, or a lead-acid battery with carbon-based electrodes. A nickel battery cell can include Ni-MH, Ni-iron, Ni-Cad, or Ni-zinc batteries. A lithium battery cell can include Li-ion, Li-ion polymer, Li-Fe-P, Li-sulfur, a nano- titanate battery, or any other type of lithium based battery cell known to those skilled in the art. The module can be comprised of battery cells of one or more battery chemistry types. [0078] The electrical connections between battery cells of a module can change the characteristics of the module. As shown in Figure 5, the module comprises battery cells that can be in a combination of parallel and serial arrangements, a parallel only arrangement or a serial only arrangement. The arrangement of battery cells in the module can be tailored based on the voltage required by the load.

[0079] Similarly, modules may have any arrangement. For example, modules can have parallel and/or serial arrangements, a serial only arrangement, or a parallel only arrangement. Batteries and/or modules may have any arrangements or connections as known in the art. See, e.g., U.S. Patent No. 5,945,806 and U.S. Patent Publication No. 2007/0252556, which are hereby incorporated by reference in their entirety. In preferable embodiments, modules may be arranged in parallel.

[0080] Since modules may be arranged in parallel, the current from a module can be much less than the required load current. This can allow smaller wires to be used to connect each module to the power bus and may reduce the cost of the electrical wiring between the module and the power bus.

[0081] In other embodiments of the invention, a capacitor may allow for storage of charge that can be accessed for power supply. Capacitors can have the ability charge quickly and provide bursts of energy when required. In some embodiments of the invention, a capacitor can be designed such that it can be used in place of a module or to augment the power supply with the ability to supply bursts of power. The capacitor connected across the power bus can also reduce high-frequency noise on the power bus associated with operating the module switches. [0082] In some embodiments, any or all of the components discussed may be provided as a single integral unit. For example, a module may be provided as a single integral unit. In some instances, one or more modules, power bus, and switches may be provided as a single integral unit. Alternatively, any or all of the components may be separable. For example, it may be easy to remove or replace a module from a battery pack system.

[0083] A power generator can be electrically connected to the battery pack with switched modules to be used as a source of power. In some embodiments, the power generator can be an electric motor, wherein the electric motor may be producing electrical power by capturing kinetic energy. The electric motor can be mechanically connected to wheels of a vehicle, wherein power generation can occur when the vehicle is braking, slowing down, or not accelerating. [0084] D. Methods of Use [0085] The method of using a battery pack with switched modules may comprise controlling the power output of the battery pack by setting the state of one or more switches that can electrically connect one or more modules to a power bus. The connectivity of the modules can be determined by monitoring the status of an individual module and setting the electrical connectivity based on calculations using equations known to those skilled in the art to produce a desired power bus voltage or current.

[0086] In other embodiments of the invention, the method of using a battery pack with switched modules can comprise controlling the charging of a module by setting the state of one or more switches that electrically connect the module to a power bus. The connectivity of the module can be determined by monitoring the state of the module or by monitoring the status of the module relative to other modules. In some embodiments of the invention, the method may further comprise charging the module using an external power source or another module as a power source.

[0087] E. Applications

[0088] The battery pack with switched modules can be used to power an electric motor, wherein the electric motor can be selected from the group consisting of: a brushless motor, a brushed motor, a coreless motor, a universal motor, and a switched reluctance machine. Electric motors can more efficiently handle a wider range of speeds and torque demands when supplied with a wider range of voltage and current. In some embodiments of the invention, the battery pack with switched modules can be used to supply a variable amount of voltage and current to an electric motor, wherein the electric motor can operate at a wide range of speeds and produce a wide range of torque. [0089] Based on a desired speed and/or torque for a motor or any other machine, the switches of the battery pack may be controlled to produce the desired power output. [0090] In one embodiment of the invention, the battery pack with switched modules can be used to power an electric motor mechanically connected to a wheel of a vehicle, where the electric motor can be an in- wheel motor, an adaptive in- wheel motor, a near wheel motor, a direct drive motor, or an electric motor mechanically connected to a wheel through a drive-train. An electric motor that is used to mechanically drive a wheel of a vehicle can be required to operate at a wide range of speeds and produce a wide range of torque. For example, a vehicle at rest can require high torque and low speed to begin moving under normal conditions, a vehicle entering a highway from a local road can require high torque and medium speed, a vehicle passing a car on the highway can require high torque and high speed, and a vehicle starting from rest on an icy road can require low speed and low torque. In some embodiments of the invention, the battery pack with switched modules can be used to supply a variable amount of voltage and current to an electric motor used in a vehicle, wherein the electric motor can operate at a wide range of speeds and produce a wide range of torque. [0091] In some embodiments of the invention, the battery with switched modules can supply a range of voltages and currents to an electric motor by selectively connecting a module to the power bus. In one embodiment of the invention, the battery with switched modules can supply a range of voltages and currents to an electric motor by pulse width modulation. In addition to PWM, another control method that may be used is current hysteresis, with control around current flowing into or out of a module. Another control method is voltage hysteresis with control around power bus voltage.

[0092] The control of the power bus voltage by selective connectivity of modules to the power bus can offer a number of advantages over only using pulse width modulation to control a supply of power to an electric motor. These advantages can include the ability to supply a greater range of voltage and current to be an electric motor, the ability to more appropriately match slowly-varying changes in power demand from an electric motor and the ability to reduce harmonic effects between the supply of power and an electric motor. [0093] II. Control of Switched Reluctance Machines

[0094] Referring to the drawings in detail, Figure 7 shows a switched reluctance machine (SRM) and how a controller, power converter, and switched reluctance motor may interact, in accordance with another aspect of the invention. A switched reluctance machine may include the controller, power converter, and switched reluctance motor. It should be understood that a switched reluctance motor might be used as a motor if the torque and speed are in the same direction and as a generator if the torque and speed are in opposing directions. A demand or operating request may be placed on the switched reluctance machine and may affect the controller. [0095] In one embodiment of the invention, a demand or operating request may originate from a source outside the SRM. For example, if the SRM were part of a vehicle, the demand may be a torque demand that may originate from a driver. In one instance, the driver of the vehicle may be pressing the accelerator in order to have the vehicle accelerate. Pressing an accelerator pedal may translate to a request for a particular torque on the motor. This translation may be linear or nonlinear. In another instance, pressing an accelerator pedal may translate to a request for a particular speed for the vehicle. The speed request from the driver action may also translate to a torque demand by comparing the requested speed to the actual speed of the vehicle and determining the torque demand. [0096] One or more sensors may be connected to a switched reluctance motor and may receive information about the motor. For example, sensors may collect information about the position of a rotor of the switched reluctance motor. Sensors could also determine the velocity of the rotor. One way of doing this may be to determine how the position of the rotor changes over time. [0097] Information from the sensors and the demand may be provided to a controller of the SRM, which may control a power converter. The controller may contain one or more look up tables (LUT) which may control the SRM to run at a given operating characteristic (such as velocity) to output a given demand (such as torque). The LUT may include an operating rule, which could be represented as an array of numbers. The operating rule may direct the actions of a power converter. [0098] A power converter may be connected to the controller and to a motor. The power converter may include a means of connecting a power source to phase windings within the motor. For example, a power converter may include a switched output converter circuit, which may have two switches per phase of the motor. For example, the operating rule could inform the controller when to switch a power converter's switches on or off based on rotor position. The motor may be controlled to respond to a torque demand. [0099] Figure 8 illustrates information flow between a controller, a power converter, and a motor. An operating request or demand may be placed on a SRM. One or more sensors may collect information from the motor and feed it to the controller. The controller may use information from the operating request and sensors to control a power converter, which may control power going to a motor, and thereby control the motor. The power going to a motor may originate from any source. In some instances, power may be provided by an energy storage system (such as a battery pack), a utility, or an energy generator. In some instances, the power source may be a variable power source. Controlling a power converter may result in controlling power provided by the power source. [00100] A controller may include one or more LUTs, which may have predetermined stored characteristics to form one or more operating rules. In one embodiment of the invention, the operating request may be a torque demand, and sensors may provide rotor position and velocity information to the controller. There may be multiple LUTs for different velocities and torque demands. For example, a LUT could be formulated for every two mph interval from zero to 78 mph, and for every ten ft- lbs of torque from zero to 190 ft-lbs, yielding eight hundred LUTs. Velocities may range from zero up to the vehicle's top speed. Torques may range from zero up to the vehicle's maximum torque. LUTs for switch states can also be calculated for situations with negative torque (such as for regenerative braking), and for situations with negative velocities (such as operating in reverse). Symmetry between LUTs, such as between operating in forward and reverse, may reduce the total number of LUTs that may be stored. [00101] Each LUT may have an operating rule that may tell the SRM when to commute the phases at a given operating characteristic to output a requested amount of mechanical power. In one embodiment of the invention, the operating rule may tell the SRM when to commute the phases, given a rotor velocity, to output a requisite torque. For example, if a vehicle were traveling 60 mph, and demanded 160 ft- lbs of torque, the controller would look for the LUT for 60 mph and 160 ft- lbs of torque. The operating rule could tell the controller how to control the power converter.

[00102] In accordance with one embodiment of the invention, the power converter may include a switched output converter circuit. A switched output converter circuit may have one or more switches per phase of the motor. In a preferable embodiment, the power converter may have two switches per motor phase. In one example, the switches may be transistor power switches that may connect to the phase windings of the motor. The operating rule could inform the controller when to switch a power converter's switches on or off based on angular rotor position. There may be more than one turn-on and one turn-off angle per revolution, in which case switches may turn on and off multiple times. The controller may compare the signal representing the angular position of the rotor with the signals representing whether the switches in a LUT should be on or off at that particular angular position and may control a power switching device so that voltage may be applied to the appropriate phase winding when the angular position of the rotor is such that a switch should be on and removed from the phase winding when the rotor's position is such that a switch should be off. [00103] In an alternate embodiment of the invention, each LUT may have multiple operating rules which may be used at different environmental conditions. For example, one such condition could be temperature. If the motor temperature were to change, its inductance may change, which may require the use of a different rule. In this embodiment, additional sensors may be attached to the SRM to evaluate which part of the LUT it could be using to operate. Another environmental condition could be the motor age, as inductance could potentially change with age. Another environmental condition could include ambient vibration. Also, environmental conditions could include desired objectives. If quieter operation was desired, a different set of LUTs that minimize torque ripple could be used.

[00104] In a preferable embodiment of the invention, the motor may be a switched reluctance motor. In an alternate embodiment of the invention, the motor may be another brushless motor. [00105] Figure 9 illustrates a method of controlling a SRM. The method may include measuring the physical properties of a switched reluctance motor 10, formulating a model for the switched reluctance motor 12, and solving the model using convex optimization algorithms 14. [00106] In one embodiment of the invention, the physical properties of a specific SRM design may be measured. The measurements may include measuring the inductance of each phase and mutual inductance between phases for varying rotor positions. For example, the measurements may take place as part of a lab test for each motor design, and may involve taking inductance measurements for each phase of the motor for a given position of the motor's rotor. Such inductance measurements may be inductance impulse response measurements. The measurements may be taken for a given interval of rotor position, such as for every degree from 0 to 360 degrees. Machine symmetry may make measurements from 0 to 360 degrees unnecessary and measurements from 0 to 120 degrees or 0 to 90 degrees may be sufficient. The measurements may yield values of inductance as a function of position. The resistance of each phase of the machine may also be measured. Other measurements of the machine's magnetic properties may also be taken, such as a magnetization hysteresis curve for each phase at a set of rotor positions.

[00107] The model for the switched reluctance motor may be formulated based on the motor's physical properties. The model may be formulated as a set of convex inequality constraint functions, affine equality constraint functions, and a convex objective function. To provide further background regarding the implementation of the embodiment, a description of model development may be set forth. [00108] In the following descriptions of model development, the following notation may be used:

[¹I represents the ceiling function, to round up to the nearest integer. z - x - * y represents element- wise multiplication, such that x and y must have the same dimension, which will result in z also having that dimension. The f^h element of z will be the product of the f¹ elements of x andy.

means x is a real-valued vector with dimension n.

means y is a vector with dimension n in which each element may take the value 1 or 2.

means A is an m-by-n real-valued matrix.

If a variable or parameter is used without specifying its dimension, it is a real-valued or integer- valued scalar. Also, subscripts do not necessarily imply an element of the vector; multiple vectors may share the same variable name but have different subscripts. [00109] In one embodiment of the invention, the voltage v at any position may either be equal to zero or equal to V_DC, the power bus voltage. In another embodiment of the invention, the voltage v at any position may also be -V_DC- This may apply when a power converter consists of switches connecting a power supply to a motor such that the switch is either open, in which case the voltage could be zero, or the switch is connected, in which case the power bus voltage may be supplied. In an alternate embodiment of the invention, the voltage at any position may include any voltage from zero to the power bus voltage. One example where voltage may vary may be when the power converter may include a type of variable voltage connector. Voltage may also vary through pulse width modulation.

[00110] Another embodiment of the invention may include another set of models which may be developed through successive approximations. A preliminary model may assume a linear current response by assuming that inductance is constant with respect to current level of a motor. The model may account for inductance variation with respect to rotor position. The model may predict the necessary switching times in order to optimally produce torque when the motor is operating at a specified angular speed. The linear current model may be defined as:

which may be minimized subject to:

And fory = \ ...m:

[00111] This final constraint ^maY ⁿ°t be convex, but can be relaxed to

In one implementation, pulse width modulation may be used to achieve any

effective voltage between 0 and V_DC- Voltage between 0 and V_DC ^may be achieved with the SRM power converter which may use pulse width modulation to achieve voltages between 0 and V_DC- A regularizing term may be added to the objective in the form of the measure of total variation of v,- from 0 and from V_DC in order to drive V₇- to either 0 or V_DC- An alternate embodiment may replace the final constraint with and relax this final

constraint to - Regularizing terms may be added to the objective to drive Vj to

either -VDC, 0, or V_Dc- [00112] The model parameters may be defined such that: where θ is a vector representing the rotor position

ω = angular speed τtot⁼ total desired torque p is defined so that the/?-norm

can be used to measure torque ripple (where typically/? =

l or 2) μ = desired trade-off weight of torque ripple relative to efficiency, such that higher// increases the relative importance of low torque ripple and reduces the relative importance of high efficiency [00113] The defined model parameter #may be a vector representing the rotor position with n evenly-spaced points, Q₁ to Q_n where Q_k = Q_max {kin). In such a case, Q_max and

n may be specified such that

[00114] In alternate embodiments of the invention, θ may represent rotor positions that may or may not be evenly spaced. [00115] The linear current model may also include parameters of a switched reluctance machine that may be measured, such as:

Tmax = the length of time required for the inductance impulse response

Δt = the smallest time interval necessary to adequately model the inductance impulse response m = the total number of phases of the SRM R = resistance in phase 7 is an inductance impulse response matrix that can map the current vector to the

voltage vector, where the k^th column of the L₁ matrix may be the discretized impulse response of the phase voltage to current, starting at the k^th element of that column and wrapping from the end of the column vector to the beginning. The impulse response is measured when the rotor is at a position of ω*k*Δt.

G = the core loss conductance of phase 7 imax ⁼ the maximum allowable root mean square current per phase ipk = the maximum allowable peak current per phase = the maximum allowable flux linkage per phase

a = the constant such that where λi is proportional to the co-energy

[00116] In a preferable embodiment of the invention, there may typically be three or four phases in an SRM, in which case m = 3 or 4. In alternate embodiments of the invention, an SRM may have different number of phases. Also, in some implementations the resistance R may be the same for all phases. In alternate implementations, R may not be the same for all phases. In some implementations, the core loss conductance G may be the same for all phases. [00117] Typically €-E «sc 1, so 1 ÷ &R *> 1. D may be a nxn first difference matrix with wrap¬

This wrap-around difference matrix may prevent

inaccuracies in the model associated with the ends of the vectors. [00118] The model may also include variables: where 1) is a vector of current through phasey^', where each element is the current at a

particular rotor position so that the Jc*¹ element is the current at position θ^ is a vector of the electrical torque produced by phasey, where each element

is the electrical torque at a particular rotor position

is a vector of the flux linkage in phasey. where V₇ is a vector of voltage across phasey, where each element is the voltage

at a particular rotor positionO η = electrical-to-mechanical efficiency σ = torque ripple

[00119] In some models, λ_j may be included for readability and may be eliminated from some models since it is entirely defined once v and i are defined. In one embodiment of the invention, as discussed previously, the voltage v at any position may either be equal to zero or equal to V_DC, the power bus voltage. In another embodiment of the invention, the voltage v at any position may also be -V_DC- This may apply when a power converter consists of switches connecting a power supply to a motor such that the switch is either open, in which case the voltage could be zero, or the switch is connected, in which case the power bus voltage may be supplied. In an alternate embodiment of the invention, the voltage at any position may include any voltage from zero to the power bus voltage. One example where voltage may vary may be when the power converter may include a type of variable voltage connector. Voltage may also be varied using pulse width modulation. [00120] The previous linear model may not account for magnetic saturation or hysteresis, which may be important physical effects that may occur in SRMs (and other magnetic machines) that may drastically alter the impulse response of both the voltage and torque to current. As current flows through a motor's stator windings, it may generate a magnetic field through the poles of the machine, which may create mechanical torque as it pulls the rotor. However, the magnetic field strength may not be linearly proportional to the current in the windings, as the linear current model assumes. The magnetic field strength may increase sub-linearly with winding current, in an effect known as saturation. Also, when the current is decreased or removed, the magnetic field may not decrease linearly, in an effect known as hysteresis. [00121] An alternate nonlinear inductance model may take magnetic saturation and hysteresis into account by altering the previous linear model. The nonlinear model may include parameters of a switched reluctance machine that may be measured, so that rather than an

inductance impulse response, may be defined to be the inductance function that

maps a current vector to a voltage vector for phasey. This function may be concave and increasing. This function could be considered the same as the inductance operator used in the previous model, except instead of each output being a fixed impulse response function (such as a fixed linear combination of inputs), each output may be a linear combination of inputs such that the linear combination may vary with the current level. This function may be modeled with a piece wise affϊne function with K pieces, such that L_jj = A_j1[I i ... i] + b_ji to L_jK = A_jK[i i ... i] + b_jK. [00122] Also, the model may be constrained such that rather than

as in the previous linear model, fory = 1...m and k = I...K, the

nonlinear model may be constrained so that

[00123] In another embodiment of the invention, a mutual inductance model may account for mutual inductance between phases in addition to magnetic saturation and hysteresis. Accounting for mutual inductance may have an effect when optimizing control. The mutual inductance model may be similar to the previous nonlinear model with all j subscripts and summations dropped, and where L may be replaced by M', where

may be the mutual inductance function. Again, this inductance function may be modeled as piecewise linear with K segments, M'_} = M_}[i i ... i]+ c; to

M' κ = M_κ[i i ... i] + c _K- Thus, the mutual inductance model objective may be defined as:

which may be minimized subject to:

where D' is a composite difference matrix. D' = diag([D D ... D]) with m total D's. [00124] This final constraint

may not be convex, but can be relaxed to . In one implementation, pulse width modulation may be used to achieve any

effective voltage between 0 and V_DC- Voltage between 0 and V_DC ^may be achieved with the SRM power converter which may use pulse width modulation to achieve voltages between 0 and V_DC- A regularizing term may be added to the objective in the form of the measure of total variation of v from 0 and from V_DC in order to drive v to either 0 or V_DC- An alternate embodiment may replace the final constraint with

^and relax this final constraint to

Regularizing terms may be added to the objective to drive v to either -VDC, 0, or VDC-

[00125] The mutual inductance model may have the same measured parameters as the previous nonlinear model, except that the matrices A and b are replaced by M and c. M and c matrices may define the piecewise-affine function used to fit the mutual inductance function, M'. M and c may include the inductance measurements from A and b in the previous nonlinear model stacked as diagonal blocks, with mutual phase-to-phase induction filling the rest of the matrix space.

[00126] The model variables for the mutual inductance model may be the same as the model variables for the previous linear and non-linear models, except that the i/s, λ_}'s, v/s and τ/s may be concatenated into vectors i, A, v, τ in Vector i may be created by stringing together the ι) vectors end-to-end to form a large vector i. Vectors A, v and τ may be similarly created by stringing together the v/s and τ/s end-to-end respectively.

[00127] The models may be solved using convex optimization. Some models may be solved using quasi-convex optimization. Such quasi-convex optimization may be done through successive iterations of convex feasibility problems. The convex optimization problem could be solved using readily-available numerical techniques. Such numerical techniques may include using programs such as MOSEK, solver.com, SeDuMi, CVX, and so forth. The solution to the problem may determine an operating rule of a LUT. In one embodiment, the solution to the convex optimization problem may determine the state of each switch within a power converter based on rotor position to optimally achieve a certain torque at a given velocity.

[00128] Figure 10 illustrates a method of controlling a switched reluctance machine for a desired optimization objective in accordance with an embodiment of the invention. The method may include measuring the physical properties of a switched reluctance motor 20, determining a desired optimization objective 22, formulating a model for the switched reluctance motor based on the measured physical properties and desired optimization objective 24, and solving the model using convex optimization 26.

[00129] In accordance with one aspect of the invention, the physical properties of a switched reluctance machine may be measured for a motor design. The measurements may include measuring the inductance of each phase and mutual inductance between phases for each rotor position. The measurements may yield values of inductance as a function of position. Other measurements may include resistance or other magnetic properties of the machine. [00130] When formulating a model for a switched reluctance motor, the model may yield an operating rule to optimally achieve a requested demand (such as torque) for a motor operating point (such as velocity). This may include determining the state of a switch within a power converter for a given rotor position to optimally yield a desired torque at the motor's operating velocity. Optimality may be defined according to determining a desired optimization objective. For example, the desired optimization objective may be efficiency, and optimality may be defined as maximizing efficiency. In another example, the desired optimization objective may be torque ripple, and optimality may be defined as minimizing torque ripple. Similarly, the desired optimization objective may be acoustic noise, and optimality may be defined as minimizing acoustic noise. The desired optimization objective may include minimizing torque delivery time. For example, torque delivery time may be how fast torque is generated by a SRM after a driver has made a torque demand (e.g. pushing the gas pedal). Desired optimization objectives may include any other characteristics of interest. The desired optimization objective may also be defined as a weighted tradeoff between two or more of these objectives, or any other characteristic of interest.

[00131] The model for the switched reluctance motor may be formulated based on the motor's measured physical properties and the desired optimization objective. The model may be formulated as a set of convex inequality constraints, affine equality constraints, and a convex objective, which may be optimized based on a desired convex optimization objective. [00132] In one embodiment of the invention the model may be defined as:

which may be minimized subject to different factors as discussed previously. The parameter μ of the model may be defined as the desired trade-off weight of torque ripple relative to efficiency, such that higher μ increases the relative importance of low torque ripple and reduces the relative importance of high efficiency. Adjusting parameter μ may be an example of adjusting a desired optimization objective. For example, giving μ a value of zero could define the desired optimization objective to be high efficiency alone.

[00133] In alternate embodiments of the invention, the model may be formulated in different ways to take different desired optimization objectives into account. The models may still be formulated as convex optimization problems, but may be formulated in a way to create parameters that can optimize for different objectives. For example, a convex optimization model may be formulated to optimize for different objectives other than efficiency and torque ripple. [00134] The model may be solved using convex optimization algorithms which may determine a global optimum for a desired optimization objective. The convex optimization problem could be solved using readily-available numerical techniques. The solution to the problem may determine an operating rule of a LUT which may optimize for the desired optimization objective. In one embodiment, the solution to the convex optimization problem may determine the state of each switch within a power converter based on rotor position to optimally achieve a certain torque demand at a given velocity, where optimality may be defined by the desired optimization objective.

[00135] Figure 11 illustrates an alternate method of controlling a switched reluctance machine. Such a method may include determining a parameter value 30, measuring SRM properties at the parameter value 32, determining a torque or velocity value 34, creating a LUT 36, determining whether there is a new torque or velocity value 38, and determining whether there is a new parameter value 40. Creating a LUT may include formulating a model for the SRM 42, and solving SRM model using convex optimization 44. When determining whether there are new torque or velocity values, if a new value is found, then the method may return to determining the torque or velocity value and creating a corresponding LUT. When determining whether there is a new parameter value, if a new value is found, the method may return to determining the parameter value and going through the corresponding following steps. Some of these steps are optional and/or may take place in different order (e.g., the method may first check whether there is a new parameter value before checking whether there is a new torque or velocity value). [00136] The method of controlling the switched reluctance machine may involve considering additional environmental conditions. Environmental conditions may include operating conditions external to the optimization model, such as the temperature of the motor, motor age, or ambient vibration level.

[00137] Motor properties may be measured in several different environmental conditions. For instance, motor properties may be measured over a set of different temperatures and may generate operating rules that can vary over temperature ranges. Since motor properties may vary at different temperatures, at a given motor temperature, measurements may be taken for motor properties such as the inductance of each phase and mutual inductance between phases for each rotor position. The measurements may yield values of inductance as a function of position at that environmental condition. [00138] Motor operating characteristics (such as velocity values) or operating requests (such as torque) may be varied when formulating a model to control an SRM. For a torque demand and a velocity value, a LUT may be created. Creating an LUT may include formulating a model considering the given torque and velocity values, the motor properties, and a desired optimization objective. The model may be formulated in a way to formulate a convex optimization problem, which may be solved using numerical convex optimization methods.

[00139] A LUT may be created for desired torque and velocity values. For example, a LUT could be formulated for every two mph interval from zero to 78 mph, and for every ten ft-lbs of torque from zero to 190 ft-lbs, yielding eight hundred LUTs. One way to create the desired LUTs may include starting with velocity and torque values of zero, creating the LUT for those values, then stepping through the velocity intervals for a given torque, and then stepping the torque value up and going through the velocity intervals and creating LUTs at each of the torque and velocity values.

[00140] A LUT may have multiple operating rules which may be used at different parameter values. For example, one parameter could include different environmental conditions, such as temperature. If the motor temperature varies, its inductance may change, which may require the use of a different rule. Since measurements of motor properties can be taken for a given parameter value, the process can be repeated for each of the parameter values or environmental conditions.

[00141] There may be symmetry within the multiple modes of operation, so a new look-up table may not need to be generated for every situation. Once the look-up tables are stored in the controller, the controller can use them, plus empirical information such as speed from a sensor and torque demand, to determine when to commute SRM phases.

[00142] In an alternate embodiment of the invention, the motor may be any brushless motor. A machine may be controlled by measuring the physical properties of a brushless motor, formulating a model for the brushless motor, and solving the model using convex optimization algorithms. The invention mentioned previously for control of switched reluctance machines may be applied to systems with brushless motors.

[00143] In another embodiment of the invention, the model may be formulated and solved as a convex optimization problem in real-time such that an embodiment would not require a LUT. Instead, a power converter may be controlled by the controller which may formulate and solve the model based on sensor readings and system demands. For instance, sensors may determine a motor operating characteristic such as motor velocity, and the system may also recognize an operating request such as torque demand. A controller may formulate the model based on the current velocity and torque demand in real time and solve the convex optimization problem to determine an operating rule for the power converter for the velocity and torque demand. [00144] It should be understood from the foregoing that, while particular implementations have been illustrated and described, various modifications can be made thereto and are contemplated herein. It is also not intended that the invention be limited by the specific examples provided within the specification. While the invention has been described with reference to the aforementioned specification, the descriptions and illustrations of the preferable embodiments herein are not meant to be construed in a limiting sense. Furthermore, it shall be understood that all aspects of the invention are not limited to the specific depictions, configurations or relative proportions set forth herein which depend upon a variety of conditions and variables. Various modifications in form and detail of the embodiments of the invention will be apparent to a person skilled in the art. It is therefore contemplated that the invention shall also cover any such modifications, variations and equivalents.

Claims

WHAT IS CLAIMED IS: 1. A battery pack with switched modules comprising: a power bus for carrying current from a variable power source, wherein the variable power source includes at least one module with one or more switches that electrically connect the module to the power bus on a selective basis, and further wherein the module contains one or more battery cells.

2. The battery pack with switched modules of claim 1 , wherein the switches include one or more diodes to control direction of current flow.

3. The battery pack with switched module of claim 1 , wherein passive electrical components possessing inductance or capacitance are used along with switches in the electrical connection to the power bus.

4. The battery pack with switched modules of claim 1 , wherein at least two modules are arranged in parallel.

5. The battery pack with switched modules of claim 1 , wherein the battery cells within each module has a combination of parallel and serial arrangements, a parallel only arrangement or a serial only arrangement.

6. The battery pack with switched modules of claim 1 , wherein power bus voltage is controlled by changing the state of one or more module switches.

7. The battery pack with switched modules of claim 4, wherein a module is intermittently connected to the power bus, and further wherein the power bus supplies continuous power by intermittent connection of another module.

8. The battery pack with switched modules of claim 3, wherein a module is intermittently connected to the power bus through a passive electrical component that has inductance.

9. The battery pack with switched modules of claim 3, wherein capacitance is added to the power bus.

10. The battery pack with switched modules of claim 1, wherein a module can be removed without disrupting power supply.

11. The battery pack with switched modules of claim 1 , wherein a module can be selectively charged by an external power supply or by another module.

12. The battery pack with switched modules of claim 1, further comprising the means to monitor the status of a module.

13. A method of controlling power flow within a battery pack, comprising: providing at least one module, wherein the module comprises one or more battery cells; and selectively connecting and/or disconnecting the module from a power bus, using one or more switches that control direction of current flow between the module and power bus.

14. A method of controlling power output of a battery pack, comprising: setting the state of one or more switches that electrically connect a module to a power bus; and monitoring the status of the module and setting the state of the switches based on the status of the module to produce a desired power bus voltage or current.

15. A method for control of a switched reluctance machine comprising: measuring physical properties of a switched reluctance machine; modeling the switched reluctance machine to formulate a convex optimization problem; and solving the convex optimization problem.

16. The method of claim 15 wherein solving the convex optimization problem determines the state of a power converter for the switched reluctance machine.

17. The method of claim 16 wherein the state of the power converter includes the positions of switches of the power converter.

18. The method of claim 15 wherein measuring a physical property of a switched reluctance machine includes measuring an inductance of a phase of the switched reluctance machine.

19. The method of claim 16 further comprising creating a look-up table with the positions of switches of the power converter.

20. A system for control of a switched reluctance machine comprising: a switched reluctance machine including a sensor generating and transmitting a sensor signal indicative of a motor operating characteristic; and a controller operatively connected to the switched reluctance machine wherein the controller controls the switched reluctance machine based on an operating request and the solution to a convex optimization problem formulated by a model of the switched reluctance machine.

21. The system of claim 20 wherein the motor operating characteristic includes the motor position or velocity.

22. The system of claim 20 wherein the operating request of a switched reluctance machine includes the desired speed or torque of the switched reluctance machine.

23. The system of claim 20 wherein the solution to the convex optimization problem determines the positions of switches of a power converter.

24. The system of claim 23 wherein the position of the switches of the power converter are stored in a look-up table.

25. The system of claim 20 wherein the convex optimization problem formulated by a model of the switched reluctance machine may be formulated or solved in real-time.

26. A method for control of a switched reluctance machine comprising: measuring a physical property of a switched reluctance machine; and controlling the switched reluctance machine based on the physical property and the solution to a convex optimization problem formulated by a model of the switched reluctance machine.

27. The method of claim 26 further comprising controlling the switched reluctance machine based on a torque or speed request.

28. The method of claim 26 wherein the convex optimization problem is formulated for a desired optimization objective.

29. The method of claim 28 wherein the desired optimization objective maximizes efficiency.

30. The method of claim 28 wherein the desired optimization objective minimizes torque ripple.

31. The method of claim 28 wherein the desired optimization objective is a trade-off between efficiency and torque ripple.

32. The method of claim 28 wherein the desired optimization objective minimizes torque delivery time.

33. A method for control of a brushless motor comprising: modeling a brushless motor to formulate a state space model; and obtaining an optimal feasible solution that satisfies a given objective using convex optimization.

34. The method of claim 33 wherein the given objective is maximizing efficiency.

35. The method of claim 33 wherein the given objective is minimizing torque ripple.

36. The method of claim 33 wherein the given objective is balancing efficiency and torque ripple.

37. The method of claim 33 wherein the given objective is minimizing torque delivery time.