Electromagnetic field generation system

Abstract

An electromagnetic field generation system and methods are provided. The system includes a plurality of parallel state predictors; and a plurality of switching components; outputs of the plurality of parallel state predictors being provided as inputs to the plurality of switching components. The method for generating an arbitrary electric current includes configuring a cost accumulator of a predictor unit to perform a cost function to obtain predicted accumulated difference of first derivative of an input variable against a reference; configuring the cost accumulator in another predictor unit to perform the cost function as a step function based on the input variable; and configuring the predictor unit to have a variable priority output as a linear function based on the input variable.

Description

[0001] DESCRIPTION

[0002] TITLE

[0003] ELECTROMAGNETIC FIELD GENERATION SYSTEM

[0004] CROSS-REFERENCE TO RELATED APPLICATION

[0005] This application claims the benefit of U.S. Provisional Application Serial No. 63 / 559,383, filed February 29, 2024, which is hereby incorporated by reference in its entirety including any tables, figures, or drawings.

[0006] BACKGROUND OF THE INVENTION

[0007] Electromagnetic field generators can be used to deliver stimulus to neurons by means of induction or direct introduction of electrical currents. By inducing electrical currents in Central Nervous System through high intensity pulses of magnetic field, a type of electromagnetic field generator, often referred to as Transcranial Magnetic Stimulation (TMS) systems, is being used as a therapeutic tool for neurological and psychiatric conditions, from treating major depression, enhancing stroke recovery, to enhancing pain relief. It can also be used as a research tool in neurosciences, particularly for the diagnosis of motor pathways.

[0008] Electromagnetic field generators for magnetic stimulation often include an externally connected electromagnetic coil. Such electromagnetic coil can be placed on the subject’s head during use, as in the case of TMS systems. The electromagnetic field is generated by injecting an electric current into the connected coil, which induces a magnetic field around the coil. When the coil is placed on the subject’s head, the generated magnetic field will further induce an electric field inside the subject’s head, which ultimately causes stimulus to the nervous system.

[0009] In order to generate stimulus by driving current across the coil, these electromagnetic field generators have various field generation circuit topologies, and their instantaneous power consumption often reaches the level of megawatts, even the duration of pulses generated are often sub-millisecond with a low repetition rate which is below 100 Hz. The loads to such electromagnetic field generators are mostly inductive, which can be characterized by the shapes of and number of turns in the coils connected. The power control to external loads in such electromagnetic field generators can be done by switching components, which are typically thyristor or silicon-controlled rectifier (SCR). The output from these typical switching components is mostly sinusoidal. Advances in transistor technologies have improved electrical characteristics of other types of switching components, e.g. insulated-gate bipolar transistor (IGBT) and metal-oxide-semiconductor field-effect transistor (MOSFET), which allow them to be used in such electromagnetic field generators for new topologies with improved controls on waveform generated, at the expense of increasing switching frequency. When generating sub-millisecond high-power arbitrary output to the load, common modulation methodologies like pulsewidth modulation (PWM) often lead to intensive switching beyond tens or hundred thousand kilohertz which results in excessive heat dissipation in switching components.

[0010] On the other hand, in high power electromagnetic field generators, parasitic inductance in system wiring often generate excessive voltage spikes during switching. In high current switching applications, the duration of a spike could be comparable to the pulse period. These spikes can be multiple times of the operating voltage of switching components, such that when repeated above the component ratings they can trigger catastrophic failures which irreversibly damage the switching components even the whole system.

[0011] One approach to suppress voltage spikes due to switching is by active clamping (ACL), which provides direct feedback of the collector potential to the gate of the switching component through TVS diodes. However, when high current is being switched repeatedly, TVS diodes can be thermally overloaded, also the feedback loop can drive the gate voltage beyond the rating of switching components. Both cases will result in system failure.

[0012] Therefore, it would be desirable to provide systems and methods capable for generating large electromagnetic field with minimized energy loss through reduced switching. In addition, it would be desirable for these systems and methods to suppress voltage spikes due to parasitic inductance safely and reliably.

[0013] BRIEF SUMMARY OF THE INVENTION

[0014] Embodiments of the subject invention pertain to an electromagnetic field generation system.

[0015] According to an embodiment of the subject invention, an electromagnetic field generation system comprises a plurality of parallel state predictors; and a plurality of switching components; wherein outputs of the plurality of parallel state predictors are provided as inputs to the plurality of switching components. Each of the plurality of switching components comprises one or more switchable high-side links. The electromagnetic field generation system may further comprise a plurality of transducers, wherein outputs of the plurality of switching components are provided as inputs to the plurality of transducers. The electromagnetic field generation system may further comprise a super-fast sensor, wherein outputs of the plurality of switching components are provided as inputs to the super-fast sensor. Moreover, outputs of the super-fast sensor are provided, with a first delay, as inputs to the plurality of parallel state predictors. The electromagnetic field generation system may further comprise a fast sensor, wherein outputs of the plurality of transducers are provided as inputs to the fast sensor. Outputs of the fast sensor are provided, with a second delay, as inputs to the plurality of parallel state predictors. The electromagnetic field generation system may further comprise a neuron, wherein outputs of the plurality of transducers are provided as inputs to the neuron. The electromagnetic field generation system may further comprise a slow sensor, wherein outputs of the neuron are provided as inputs to the slow sensor. The electromagnetic field generation system may further comprise a filter, wherein outputs of the slow sensor are provided, with a third delay, as inputs to the filter. Outputs of the filter are provided as inputs to the plurality of parallel state predictors. Further, the first delay, the second delay, and the third delay are different from one another.

[0016] According to another embodiment of the subject invention, a method for generating an arbitrary electric current is provided. The method comprises configuring a cost accumulator of a predictor unit to perform a cost function to obtain predicted accumulated difference of first derivative of an input variable against a reference; configuring the cost accumulator in another predictor unit to perform the cost function as a step function based on the input variable; and configuring the predictor unit to have a variable priority output as a linear function based on the input variable.

[0017] According to another embodiment of the subject invention, a method for rapid adjustment of output voltages from switching components is provided. The method comprises configuring a state screener to include a lookup table for charging direction in a half-bridge module in corresponding states; configuring a predictor unit to have a variable priority output as a linear function of a first derivative of electric currents of transducers; and configuring a handler which contains a lookup table for alternative state which rapidly dissipates stored energy of capacitors and / or load via resistive elements. According to another embodiment of the subject invention, a method for suppressing local voltage surge in switching devices within switching components is provided. The method comprises configuring a predictor unit to have an instruction output based on a previous state output stored in a same parallel state predictor; and configuring the predictor unit to have a variable priority output as a linear function of difference between an input variable and a reference.

[0018] According to another embodiment of the subject invention, a method of handling asynchronous delayed feedback signals is provided. The method comprises configuring a compensator in a parallel state predictor to predict a real-time value of a feedback signal as a linear function of a variable, by interpolating the previous values of another signal stored in cache of the same parallel state predictor.

[0019] According to another embodiment of the subject invention, a non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method for generating an arbitrary electric current is provided. The method comprises configuring a cost accumulator of a predictor unit to perform a cost function including the predicted accumulated difference between a first derivative of an input variable and a reference; configuring the cost accumulator in another predictor unit to perform a cost function as a step function based on the input variable; and configuring the predictor unit to have a variable priority output as a linear function based on the input variable.

[0020] According to another embodiment of the subject invention, a non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method for rapid adjustment of output voltage from switching components is provided. The method comprises configuring a state screener to include a lookup table for charging direction in a half-bridge module in corresponding states; configuring a predictor unit to have a variable priority output as a linear function of a first derivative of the electric currents of transducer; and configuring a handler comprising a lookup table for alternative state which rapidly dissipates stored energy of capacitors and / or load via resistive elements.

[0021] According to another embodiment of the subject invention, a non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method for suppressing local voltage surge in switching devices within switching components is provided. The method comprises configuring a predictor unit to have an instruction output based on a previous state output stored in a same parallel state predictor; and configuring the predictor unit to have a variable priority output as a linear function of difference between an input variable and a reference.

[0022] According to another embodiment of the subject invention, a non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method of handling asynchronous delayed feedback signals is provided. The method comprises configuring a compensator in a parallel state predictor to predict a real-time value of a feedback signal as a linear function of a variable, by interpolating previous values of another signal stored in cache of the same parallel state predictor.

[0023] BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIG. 1 is a system overview illustrating a block diagram of an electromagnetic field generation system, according to an embodiment of the subject invention.

[0025] FIG. 2 is an overview of Parallel State Predictor (PSP) illustrating the block diagram of a parallel state predictor, according to an embodiment of the subject invention.

[0026] FIG. 3 is an example layout of switching component illustrating a schematic diagram of a set of switching component, which forms a multilevel inverter, according to an embodiment of the subject invention.

[0027] FIGS. 4A-4C show simulated flexible step predictive control illustrating the simulated system response under a PSP with single predictor unit bound to electric current in the transducers, wherein the reference signal is a sine wave with a period of 2 ms, wherein FIG. 4A illustrates the predicted responses from all cost accumulators in PSP over the first 2 steps, wherein the identifiers of cost accumulators are tagged at the end of each trace respectively, wherein the corresponding gate combination of each identifier is shown in Table 1, wherein FIG. 4B illustrates the predicted responses from all cost accumulators in PSP over all steps covering one full period of a repeating sinusoidal reference signal, and wherein FIG. 4C illustrates the predicted responses from the optimal control output from PSP (solid line) against reference signal (dashed line), according to an embodiment of the subject invention.

[0028] TABLE 1. Corresponding gate combination and output voltage to state identifiers in the simulation of flexible step predictive control.

[0029] DETAILED DISCLOSURE OF THE INVENTION

[0030] According to the embodiments of the subject invention, an electromagnetic field generation system and methods for therapeutic neural modulation, and non-invasive bidirectional brain-machine interface (BMI) are provided.

[0031] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term “and / or” includes any and all combinations of one or more of the associated listed items. As used herein, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well as the singular forms, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises” and / or “comprising,” when used in this specification, specify the presence of stated features, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof.

[0032] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one having ordinary skill in the art to which this invention pertains. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and the present disclosure and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

[0033] When the term “about” is used herein, in conjunction with a numerical value, it is understood that the value can be in a range of 90% of the value to 110% of the value, i.e. the value can be + / - 10% of the stated value. For example, “about 1 kg” means from 0.90 kg to 1.1 kg.

[0034] System Composition

[0035] The minimal set of an electromagnetic field generation system comprises one or more parallel state predictor(s) 10 and switching component(s) 20, as shown in Figure 1. The parallel state predictors 10 act as the controller for the power modulation in the switching components 20. Moreover, the output of the switching components 20 may be connected to one or more transducers 30.

[0036] In one of the embodiments, the transducers 30 are magnetic coils.

[0037] The system may additionally comprise one or more sensors 41-43 to provide feedback information to the parallel state predictors 10.

[0038] In an embodiment, the system may comprise voltage sensors with bandwidth of 10MHz, which are denoted as “super fast sensor” 41 in Figure 1, sensing the output voltage from the switching components 20.

[0039] In another embodiment, the system may comprise hall elements with bandwidth of 1MHz, which are denoted as “fast sensor” 42 in Figure 1, sensing the magnetic field from the transducers 30.

[0040] In another embodiment, the transducers 30 induce an electric current on neurons 60, which further modulates the postsynaptic potentials. The system may comprise patch electrodes with bandwidth of in the order of 10kHz or below, which are denoted as “slow sensor” 43 in Figure 1, sensing the bio-electric signals or electroencephalographic (EEG) signals containing the induced signals.

[0041] All three embodiments above are subject to various levels of delays, including but not limited to, delays due to sampling in measurement or propagation of signals and filtering, before feeding into parallel state predictors. The delays are denoted as z~A, z~B, and z~cfor three different groups of sensors, namely, the group of the slow sensors 41, the group of the fast sensors 42, and the group of the super fast sensors 43 as shown in Figure 1. In one embodiment, the three groups of sensors 41-43 may have different delays z~A, z~B, and z~c, respectively. In particular, the delays from sensors within the same group of sensors are the same, as shown in Figure 1; while the delays across different groups of the sensors may be different and the ranges of the three delays z~A, z~B, and z~ccan overlap.

[0042] In another embodiment, the system may comprise one or more groups of sensors. Specifically, the system may include any combination of the slow sensors 41, fast sensors 42, and super-fast sensors 43. Furthermore, the system may have more than one groups of sensors operating at the same speed, while their delays may vary from one another.

[0043] Furthermore, sensors in all embodiments may subject to noises, disturbances and artifacts in the sensor input. Disturbances may be caused by accidental movement of transducers 30, also noises can be picked up through electromagnetic coupling, while artifacts introduced by nearby signal sources. These are denoted as disturbances in Figure 1.

[0044] The signal acquired from sensor may contain significant artifacts which requires a filter before feeding into parallel state predictors. In an embodiment, muscle artifacts are picked up by slow sensors 43 as shown in Figure 1.

[0045] In an embodiment, the system comprises a filter 50 comprising a Fourier transformer and a processor, which feed spectral intensities of input signal within the band of 100 Hz - 1000 kHz at the interval of 100 Hz to the parallel state predictor 10.

[0046] It is noted that the filter 50 can also be any data processor unit which removes noises, disturbances or artifacts from sensor signals. For example, the filter 50 is an independent component analysis (ICA) processor.

[0047] It is noted that the description above is not exhaustive. The electromagnetic field generation may contain no or any number of sensors, which does not impact the generality of the system description.

[0048] Further descriptions of parallel state predictors 10 and switching components 20 are given below.

[0049] Controller and Methods

[0050] The controller provides driving signals to control switching in half-bridge modules, which include high-side link, and charger connections. It is implemented by one or more parallel state predictors (PSP) 10. Several control methods, including parallel state prediction methods and adaptive charging control methods are implemented in PSP to generate the driving signals and simultaneously update the controller model parameters. Parallel State Predictor (PSP)

[0051] A conventional method for generating arbitrary output in electromagnetic field generation system involves pulsewidth modulation (PWM) with low-pass filtering. However, switching frequency for PWM could only reach 100kHz in maximum with available switching components, inducing significant switching loss at high frequencies while unable to track reference (~10kHz) in lower switching frequencies. It also requires inductive filter to be fitted before the load at output, which could to bulky due to high current ratings required by load.

[0052] According to the embodiments of the subject invention, the parallel state predictor (PSP) 10, based on different approaches, is utilized to generate control signals for arbitrary output which lower switching frequency in average. An overview of the components in parallel state predictor is shown in Figure 2.

[0053] In an embodiment, the parallel state predictor 100 comprises one or more compensators 110, one or more predictor units 130, a parameter updater 120, and a handler 140

[0054] The compensator 110 is configured to take delayed sensor feedback and predicts compensated current feedback for the time delay and lumped modeling error, ensuring that all feedback signals, regardless of delay durations, are evaluated from the same reference time across predictor units.

[0055] The predictor unit 130 is configured to receive time-compensated feedback of the system and generates instructions to the handler.

[0056] In an embodiment, the instructions include a numerical priority value and a set of one or more gate combinations for switching components, denoted by “state” as shown in Table 1.

[0057] Herein, Table 1 compares output voltage for various capacitor charge scenarios over different gate combinations (“state”) in a system with switching components comprising 4 cascaded half-bridge modules.

[0058] Subscripts in V denotes the voltage in capacitor of module A, B, C and D respectively. For example, V10-10-10-10 refers to the MLI output when capacitor voltages in all modules are set to IkV (10 x 100 V).

[0059] Specifically, the predictor unit 130 is responsible for calculating predicted system response over a finite time step for one or more gate combinations in switching components (state), which is performed in parallel by cost accumulators 131. A predictor unit 130 can contain one or more cost accumulators 131 and the predictor units inside a PSP 100 can be asynchronous.

[0060] In an embodiment, the output of a predictor unit (“instruction”) includes a priority value, which is a positive integer, and a set containing one or more gate combinations in switching components (states).

[0061] In an embodiment, a cost accumulator 131 is configured to calculate the predicted system response for one particular state with propagated information.

[0062] In another embodiment, a cost accumulator 131 is configured to calculate the cost of one particular state with propagated information. Such cost may be the difference between the system state derived from the feedback information and the preset reference as set in a cost accumulator by hardwired constant or propagated variable.

[0063] In an embodiment, the total number of states in system is more than the number of cost accumulators 131 in a predictor unit 130. Only a subset of state which is likely to contain the optimal state is calculated in predictor unit 130, which is selected by a state screener 121 in a parameter updater 120. The parameter updater 120 is configured to calculate the model parameters in predictor unit 130 based on feedback update and cache, and the results are propagated to predictor unit 130 periodically to avoid drifting or accumulation of error.

[0064] In one or more embodiments, the predictor unit 130 may include a type A cost accumulator configured to perform a cost function to obtain predicted accumulated difference of first derivative of an input variable against a reference. Alternatively, the predictor unit 130 may have a type B cost accumulator configured to perform the cost function as a step function based on the input variable.

[0065] In certain embodiment, the predictor unit 130 of the PSP 100 may comprise any combination of type A, type B, or other suitable kind of cost accumulators 131. In particular, the system may comprise either one type of the cost accumulators 131 or more than one types of the cost accumulators 131.

[0066] In one embodiment, as shown in Figure 2, a priority module 132 may be optionally added to the predictor unit 130 to perform a bypass method. The priority module 132 can be configured to have a variable output as a linear function based on the input variable. For instance, the output of the predictor unit 130 originated from its priority module 132 generally has a higher priority than the output originated from its cost accumulator 131. It is noted that the priority module 132 is optional within the predictor unit 130. In other words, the predictor unit 130 of the PSP 100 may comprise the priority module 132 or may function without the priority module 132.

[0067] Further, the handler 140 receives instructions from predictor unit and updates the optimal gate combination for switching components continuously. In an embodiment, when a handler 140 updates its output, it passes its existing output to a cache 150, as shown in Figure 2.

[0068] In another embodiment, the previously compensated sensor feedbacks and gate combination outputs are stored in the cache 150.

[0069] In an embodiment, in case energy needs to be taken off the coil in emergency without generating any significant stimulus due to instantaneous change of current (dl / dt), lookup table for alternative state to be deployed under specific state are stored in the handler.

[0070] In an embodiment, the one or more parallel state predictors 10 are implemented by a field-programmable gate array (FPGA).

[0071] In another embodiment, the one or more parallel state predictors 10 are implemented by a system comprising one or more microprocessor and flash memory or random-access memory (RAM).

[0072] It is noted that the implementation methods of PSP in hardware are not exhaustive, and could be used in combination.

[0073] In an embodiment, the hardware implementation of PSP includes a combination of microprocessor and FPGA applying PSP on a hybrid System-on-Chip (Xilinx Zynq SoC), which predictor units in PSP are implemented on FPGA and other PSP components implemented by sequential logic with a microprocessor (ARM Cortex-A9).

[0074] Switching Components

[0075] In an embodiment, the switching components 20 contain one or more half-bridge modules, which are cascaded when there are more than one half-bridge modules. Each halfbridge module contains two switching components and a capacitor bank.

[0076] In another embodiment, the switching components 20 in the electromagnetic field generation system comprise four cascaded half-bridge modules shown by dashed rectangles with 8 output levels, as shown in Figure 3. In addition, each cascading half-bridge module comprises two switching devices and a capacitor bank. According to the embodiments of the subject invention, the switching devices may be any devices with switching characteristics, including but not limited to, MOSFET, IGBT, SCR and thyristor. It is noted that the number of cascading half-bridges in the figure is not exhaustive and could be any number.

[0077] In another embodiment, the switching components 20 comprise hardwired or switched high-side links between a pair of half-bridges, as denoted by dotted lines in Figure 3. High-side link can be any inductive and capacitive components to facilitate energy storage in and transfer between capacitors during conduction (For example, suppress voltage spike in high-side link).

[0078] Each capacitor bank may comprise one or more capacitors in series, in parallel or in hybrid configurations.

[0079] In an embodiment, the capacitance in each capacitor bank is made inequal to facilitate output of complex waveforms containing multiple frequencies, output of lower-amplitude, or dynamic waveforms at a much higher frequency compared to the carrier wave.

[0080] Energy Charging in Switching Components

[0081] Due to vast instantaneous load, the energy flowing through half-bridge modules are mainly supplied by the capacitors in module. The energy in capacitors is lost, when powering load or dissipating over time and is replenished by chargers.

[0082] Figure 3 illustrates the system layout under State #10 at the 30-20-10-10 capacitor configuration listed in Table 1. “1” in gate combination refers to closed high-side switch and opened low-side switch in a module, while “0” refers otherwise.

[0083] In an embodiment, a charger comprises a boost circuitry to raise mains voltage (for example, 110 or 220V) to target capacitor voltage as denoted next to capacitor symbol as shown in Figure 3. The charger only recharges capacitors in one half-bridge module at a time, as the half-bridge modules are isolated to each other.

[0084] In another embodiment, a charger is connected to one half-bridge module by a switch circuitry. By closing high-side link (denoted by dotted line in Figure 3) between the charger connected module and another half-bridge module, capacitors in both modules may be charged simultaneously.

[0085] In another embodiment, more than one charger is connected to switching components to facilitate charging rate in single half-bridge module when connected in parallel. In another embodiment, more than one charge is connected to switching components to facilitate charging rate in separate half-bridge modules when connected independently.

[0086] In an embodiment, chargers are configured to charge in according to a fixed or variable voltage reference (for example, fixed at 1400V or between 1200 and 1600V).

[0087] In another embodiment, chargers are configured to deliver a set amount of charge without reference to voltage (for example, delivering 100 J over 100 ms). The electric current can be fixed or variable in the process to reduce unnecessary control effort.

[0088] Referring to Figures 4A-4C, simulated flexible step predictive control illustrates the simulated system response under a PSP with single predictor unit bound to electric current in the transducers, wherein the reference signal is a sine wave with a period of 2 ms. In particular, Figure 4A illustrates the predicted responses from all cost accumulators in PSP over the first 2 steps, wherein the identifiers of cost accumulators are tagged at the end of each trace respectively, wherein the corresponding gate combination of each identifier is shown in Table 1. Figure 4B illustrates the predicted responses from all cost accumulators in PSP over all steps covering one full period of a repeating sinusoidal reference signal. Figure 4C illustrates the predicted responses from the optimal control output from PSP (solid line) against reference signal (dashed line).

[0089] According to the embodiments of the invention, an electromagnetic field generation system is provided, comprising a parallel state predictor as a controller, a plurality of switching components, and a plurality of transducers. Moreover, methods for configuring the electromagnetic field generation system through parallel state predictor are provided, for surge limitation and optimization of tracking for arbitrary outputs.

[0090] Embodiment 1. An electromagnetic field generation system, comprising: a plurality of parallel state predictors; and a plurality of switching components; wherein outputs of the plurality of parallel state predictors are provided as inputs to the plurality of switching components.

[0091] Embodiment 2. The electromagnetic field generation system of embodiment 1, wherein each of the plurality of switching components comprises one or more switchable high-side links. Embodiment 3. The electromagnetic field generation system of embodiment 1 or 2, further comprising a plurality of transducers, wherein outputs of the plurality of switching components are provided as inputs to the plurality of transducers.

[0092] Embodiment 4. The electromagnetic field generation system of any preceding embodiment, further comprising a super-fast sensor, wherein outputs of the plurality of switching components are provided as inputs to the super-fast sensor.

[0093] Embodiment 5. The electromagnetic field generation system of any preceding embodiment, wherein outputs of the super-fast sensor are provided, with a first delay, as inputs to the plurality of parallel state predictors.

[0094] Embodiment 6. The electromagnetic field generation system of any preceding embodiment, further comprising a fast sensor, wherein outputs of the plurality of transducers are provided as inputs to the fast sensor.

[0095] Embodiment 7. The electromagnetic field generation system of any preceding embodiment, wherein outputs of the fast sensor are provided, with a second delay, as inputs to the plurality of parallel state predictors.

[0096] Embodiment 8. The electromagnetic field generation system of any preceding embodiment, further comprising a neuron, wherein outputs of the plurality of transducers are provided as inputs to the neuron.

[0097] Embodiment 9. The electromagnetic field generation system of any preceding embodiment, further comprising a slow sensor, wherein outputs of the neuron are provided as inputs to the slow sensor.

[0098] Embodiment 10. The electromagnetic field generation system of any preceding embodiment, further comprising a filter, wherein outputs of the slow sensor are provided, with a third delay, as inputs to the filter. Embodiment 11. The electromagnetic field generation system of any preceding embodiment, wherein outputs of the filter are provided as inputs to the plurality of parallel state predictors.

[0099] Embodiment 12. The electromagnetic field generation system of any preceding embodiment, wherein the first delay, the second delay, and the third delay are different from one another.

[0100] Embodiment 13. A method for generating an arbitrary electric current, the method comprising: configuring a cost accumulator of a predictor unit to perform a cost function to obtain predicted accumulated difference of first derivative of an input variable against a reference.

[0101] Embodiment 14. A method for generating an arbitrary electric current, the method comprising: configuring a cost accumulator of a predictor unit to perform the cost function as a step function based on the input variable.

[0102] Embodiment 15. A method for generating an arbitrary electric current, the method comprising: configuring a first cost accumulator of a first predictor unit to perform a cost function to obtain predicted accumulated difference of first derivative of an input variable against a reference; configuring a second cost accumulator of a second predictor unit to perform the cost function as a step function based on the input variable; and configuring one or both of the first and second predictor units to have a variable priority output as a linear function based on the input variable.

[0103] Embodiment 16. A method for rapid adjustment of output voltages from switching components, the method comprising: configuring a state screener to include a lookup table for charging direction in a half-bridge module in corresponding states; configuring a predictor unit to have a variable priority output as a linear function of a first derivative of electric currents of transducers; and configuring a handler which contains a lookup table for alternative state which rapidly dissipates stored energy of capacitors and / or load via resistive elements.

[0104] Embodiment 17. A method for suppressing local voltage surge in switching devices within switching components, the method comprising: configuring a predictor unit to have an instruction output based on a previous state output stored in a same parallel state predictor; and configuring the predictor unit to have a variable priority output as a linear function of difference between an input variable and a reference.

[0105] Embodiment 18. A method of handling asynchronous delayed feedback signals, comprising: configuring a compensator in a parallel state predictor to predict a real-time value of a feedback signal as a linear function of a variable, by interpolating the previous values of another signal stored in cache of the same parallel state predictor.

[0106] Embodiment 19. A non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method for generating an arbitrary electric current, the method comprising: configuring a cost accumulator of a predictor unit to perform a cost function including the predicted accumulated difference between a first derivative of an input variable and a reference, configuring the cost accumulator in another predictor unit to perform a cost function as a step function based on the input variable; and configuring the predictor unit to have a variable priority output as a linear function based on the input variable.

[0107] Embodiment 20. A non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method for rapid adjustment of output voltage from switching components, the method comprising: configuring a state screener to include a lookup table for charging direction in a half-bridge module in corresponding states; configuring a predictor unit to have a variable priority output as a linear function of a first derivative of the electric currents of transducer; and configuring a handler comprising a lookup table for alternative state which rapidly dissipates stored energy of capacitors and / or load via resistive elements.

[0108] Embodiment 21. A non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method for suppressing local voltage surge in switching devices within switching components, the method comprising: configuring a predictor unit to have an instruction output based on a previous state output stored in a same parallel state predictor; and configuring the predictor unit to have a variable priority output as a linear function of difference between an input variable and a reference.

[0109] Embodiment 22. A non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method of handling asynchronous delayed feedback signals, the method comprising: configuring a compensator in a parallel state predictor to predict a real-time value of a feedback signal as a linear function of a variable, by interpolating previous values of another signal stored in cache of the same parallel state predictor.

[0110] All patents, patent applications, provisional applications, and publications referred to or cited herein are incorporated by reference in their entirety, including all figures and tables, to the extent they are not inconsistent with the explicit teachings of this specification.

[0111] It should be understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in light thereof will be suggested to persons skilled in the art and are to be included within the spirit and purview of this application and the scope of the appended embodiments. In addition, any elements or limitations of any invention or embodiment thereof disclosed herein can be combined with any and / or all other elements or limitations (individually or in any combination) or any other invention or embodiment thereof disclosed herein, and all such combinations are contemplated with the scope of the invention without limitation thereto.

[0112] REFERENCES

[0113] US10074997B2 - Method and apparatus for creating a dynamically reconfigurable energy storage device US10950378B2 - Methods and systems for controlling electromagnetic field generators

[0114] JP2009536073A - Deep brain localization transtemporal magnetic stimulation system based on trajectory

Claims

CLAIMSWe claim:

1. An electromagnetic field generation system, comprising: a plurality of parallel state predictors; and a plurality of switching components; wherein outputs of the plurality of parallel state predictors are provided as inputs to the plurality of switching components.

2. The electromagnetic field generation system of claim 1, wherein each of the plurality of switching components comprises one or more switchable high-side links.

3. The electromagnetic field generation system of claim 1, further comprising a plurality of transducers, wherein outputs of the plurality of switching components are provided as inputs to the plurality of transducers.

4. The electromagnetic field generation system of claim 3, further comprising a superfast sensor, wherein outputs of the plurality of switching components are provided as inputs to the super-fast sensor.

5. The electromagnetic field generation system of claim 4, wherein outputs of the super-fast sensor are provided, with a first delay, as inputs to the plurality of parallel state predictors.

6. The electromagnetic field generation system of claim 5, further comprising a fast sensor, wherein outputs of the plurality of transducers are provided as inputs to the fast sensor.

7. The electromagnetic field generation system of claim 6, wherein outputs of the fast sensor are provided, with a second delay, as inputs to the plurality of parallel state predictors.

8. The electromagnetic field generation system of claim 7, further comprising a neuron, wherein outputs of the plurality of transducers are provided as inputs to the neuron.

9. The electromagnetic field generation system of claim 8, further comprising a slow sensor, wherein outputs of the neuron are provided as inputs to the slow sensor.

10. The electromagnetic field generation system of claim 9, further comprising a filter, wherein outputs of the slow sensor are provided, with a third delay, as inputs to the filter.

11. The electromagnetic field generation system of claim 10, wherein outputs of the filter are provided as inputs to the plurality of parallel state predictors.

12. The electromagnetic field generation system of claim 11, wherein the first delay, the second delay, and the third delay are different from one another.

13. A method for generating an arbitrary electric current, the method comprising: configuring a cost accumulator of a predictor unit to perform a cost function to obtain predicted accumulated difference of first derivative of an input variable against a reference.

14. A method for generating an arbitrary electric current, the method comprising: configuring a cost accumulator of a predictor unit to perform the cost function as a step function based on the input variable.

15. A method for generating an arbitrary electric current, the method comprising: configuring a first cost accumulator of a first predictor unit to perform a cost function to obtain predicted accumulated difference of first derivative of an input variable against a reference; configuring a second cost accumulator of a second predictor unit to perform the cost function as a step function based on the input variable; and configuring one or both of the first and second predictor units to have a variable priority output as a linear function based on the input variable.

16. A method for rapid adjustment of output voltages from switching components, the method comprising: configuring a state screener to include a lookup table for charging direction in a half-bridge module in corresponding states; configuring a predictor unit to have a variable priority output as a linear function of a first derivative of electric currents of transducers; and configuring a handler which contains a lookup table for alternative state which rapidly dissipates stored energy of capacitors and / or load via resistive elements.

17. A method for suppressing local voltage surge in switching devices within switching components, the method comprising: configuring a predictor unit to have an instruction output based on a previous state output stored in a same parallel state predictor; and configuring the predictor unit to have a variable priority output as a linear function of difference between an input variable and a reference.

18. A method of handling asynchronous delayed feedback signals, comprising: configuring a compensator in a parallel state predictor to predict a real-time value of a feedback signal as a linear function of a variable, by interpolating the previous values of another signal stored in cache of the same parallel state predictor.

19. A non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method for generating an arbitrary electric current, the method comprising: configuring a cost accumulator of a predictor unit to perform a cost function including the predicted accumulated difference between a first derivative of an input variable and a reference, configuring the cost accumulator in another predictor unit to perform a cost function as a step function based on the input variable; and configuring the predictor unit to have a variable priority output as a linear function based on the input variable.

20. A non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method for rapid adjustment of output voltage from switching components, the method comprising: configuring a state screener to include a lookup table for charging direction in a half-bridge module in corresponding states; configuring a predictor unit to have a variable priority output as a linear function of a first derivative of the electric currents of transducer; and configuring a handler comprising a lookup table for alternative state which rapidly dissipates stored energy of capacitors and / or load via resistive elements.

21. A non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method for suppressing local voltage surge in switching devices within switching components, the method comprising: configuring a predictor unit to have an instruction output based on a previous state output stored in a same parallel state predictor; and configuring the predictor unit to have a variable priority output as a linear function of difference between an input variable and a reference.

22. A non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform a method of handling asynchronous delayed feedback signals, the method comprising: configuring a compensator in a parallel state predictor to predict a real-time value of a feedback signal as a linear function of a variable, by interpolating previous values of another signal stored in cache of the same parallel state predictor.

Images