Systems and methods for multi-converter maximum power point tracking

Abstract

An architecture of an integrated multi-converter, direct-drive, maximum power point tracker (MPPT) is provided. The MPPT can be used to power a plurality of non-linearly related loads via dual buck converters that can simultaneously optimize the powers to the loads for maximum performance. In some embodiments, the MPPT of the present embodiments can command multiple power converters, to power a direct-drive, photovoltaic electrodialysis (ED) desalination system while still tracking the maximum power point. The MPPT can be integrated with downstream converters to eliminate use of an entire power converter and energy storage system, reducing cost and increasing overall system efficiency.

Description

Attorney Docket No.: MIT 25641 USPCT | 88212-431989SYSTEMS AND METHODS FOR MULTI-CONVERTER MAXIMUM POWER POINT TRACKINGCROSS REFERENCE TO RELATED APPLICATION(S)

[0001] The present disclosure claims priority to and the benefit of U.S. Provisional Application No. 63 / 729,464, entitled "System Architectures of Integrated Multi-Converter, Direct-Drive Maximum Power Point Trackers," filed on December 8. 2024, the content of which is incorporated by reference herein in its entirety.FIELD

[0002] The present disclosure relates to systems and methods for multi-converter maximum power point tracking for operating loads from photovoltaic arrays, and more particularly relates to systems and methods for direct-drive maximum power point trackers that use integrated multi-converter architectures to provide power to multiple downstream loads without intermediate energy storage.BACKGROUND

[0003] Renewable desalination without energy storage is a grand challenge. Energy’ storage can increase the capital and operating expenditure, logistical issues with shipping and procurement, and complexity of system operation, while having environmental implications. In fact, energy' storage is a significant driver of cost, maintenance, and complexity when integrating renewables with desalination.

[0004] Photovoltaic (PV) systems have been shown to affordably desalinate water for resource-constrained communities. Previous works have shown that electrodialysis (ED) systems can operate off-grid cost-effectively and at higher water recovery' ratios (ratio of clean product water to feed water) for small village scale production volumes than reverse osmosis systems, for example. Solar PV energy is projected to grow from about 3% of the energy supply of the United States to about 45% by 2050 and is the renewable energy technology w ith the highest growth rate. How ever, solar photovoltaic panels present a challenge of intermittent energy supply due at least to fluctuations in solar irradiance. These intermittencies can cause issues for loads that are typically run at constant power. They also create challenges in drawing the maximum pow er from the solar panels.Attorney Docket No.: MIT 25641 USPCT | 88212-431989

[0005] Specifically, as the solar irradiance fluctuates, at times rapidly in the case of clouds, the maximum power available may change. Photovoltaic panels have current-voltage (IV) curves that are characterized by a relatively constant short-circuit panel current that drops rapidly close to the open circuit panel voltage. As solar irradiance increases during the day, the incident sunlight increases the amount of available current, while the panel remains voltage limited by its open circuit panel voltage. Even early on in the day, with limited sunlight, the panel will reach its open circuit voltage quickly, but will be limited in that very little current can be rawn before rapidly dropping the panel voltage. Moreover, aside from the variations in IV curves due to irradiance, there can be variations in panel behavior due to temperature. For example, decreasing the temperature of the photovoltaic array can increase the open circuit voltage, and thus the power available for a panel of equivalent available current. Conversely, increasing the temperature increases losses via resistance in the system and in the crystal lattice of the doped silicon, which explains why many panels on larger or more complex arrays may be actively or passively cooled.

[0006] Further still, the solar panel array configuration may affect the characteristic IV curve seen downstream of the system. For example, an array in series increases the open circuit voltage, but does not affect the short circuit current, while an array in parallel increases the short circuit current, but does not increase the open circuit voltage. This translates to a shifting of the IV curve upward (parallel) or to the right (series). Other effects that affect IV curves can include mismatch losses, such as partial shading of solar arrays, soiling of the array, and damaged cells. These are particularly challenging in that they create multiple local optima for maximum power. Additional characteristic features of the IV curve for a PV panel can be caused by series and shunt resistances. Series resistances can decrease the trailing edge slope, and shunt resistances can increase the leading edge slope.

[0007] Accordingly, there is a need for systems and methods for cost-effective management of solar energy.SUMMARY

[0008] The present application is directed to an integrated maximum power point tracker (MPPT) photovoltaic (PV) electrodialysis (ED) system that can enable desalination without battery storage. Typical MPPTs can use one power converter (e.g., a buck-boost) to draw the maximum power from a solar array by algorithmically altering the duty' cycle to charge aAttorney Docket No.: MIT 25641 USPCT | 88212-431989 battery, which then supplies power to multiple loads downstream. However, some applications use two or more loads directly-driven by the photovoltaic array, rather than buffered by a battery or other energy storage. The MPPT of the present embodiments can use an integrated dual buck converter with an MPPT, with a plurality of the buck converters simultaneously optimizing load balancing for maximum water production within an ED stack while tracking the maximum power point. A nested feedback loop, which uses a continuously adjusted weight on one converter's duty cycle, can be implemented, with the rest of the control being algorithmically similar to canonical voltage setpoint tracking and perturb and observe methods. It wi 11 be appreciated that while the present disclosure discusses use of the MPPT with an ED stack, the MPPT of the present embodiments can be used in combination with a variety of systems, such as household appliances, machinery, aeronautics, water processing systems, filtration systems, waste processing systems, management of dual or multi -pumping architectures, solar HVAC systems (e.g., chillers, hot water heaters), refrigeration and freezing units, and / or charging system for fleets of electric vehicles.

[0009] In an aspect, embodiments relate to a direct-drive system architecture comprising a photovoltaic (PV) array. A maximum power point tracker (MPPT) is in communication with the PV array. A plurality of power converters are located downstream of the MPPT and in communication with the MPPT and the PV array, with the plurality of power converters being configured to receive a duty cycle command from the MPPT and at least one of a voltage and a current from the PV array. A plurality of loads are provided, with each load being downstream of, and in communication with, one or more of the plurality of power converters, with each of the plurality' of power converters being configured to independently regulate one or more of the plurality of loads. The MPPT is configured to regulate: (i) each of the downstream plurality of loads independently; and (ii) a power point of the PV array.

[0010] One or more of the following features can be included. The plurality of power converters can be variable direct current (DC) / DC converters. The plurality of loads and the power point of the PV array can be regulated simultaneously. The system architecture can further comprise a feedback loop that includes a weighting factor on a duty cycle on at least one of the plurality of power converters. The weighting factor can be dynamically adjusted based on sensed process variables from the plurality of loads. The plurality of power converters can comprise a dual buck converter. The plurality of power converters can operate in a continuous conduction mode with a current ripple constraint or a discontinuousAttorney Docket No.: MIT 25641 USPCT | 88212-431989 conduction mode. The MPPT can comprise a plurality of power supplies, with each power supply being in communication with a power converter of the plurality of power converters. Each load associated with each of the lurality of power converters can be configured to be operated simultaneously. The plurality of loads can include a hydraulic load and an applied electrical load. The plurality of loads can include an electrodialysis (ED) stack or a pump. No converter can exist upstream of the MPPT. The PV array and the MPPT can be not in communication with a battery. The plurality of loads can comprise one or more of water processing systems, filtration systems, waste processing systems, management of dual or multi-pumping architectures, solar ETVAC systems, refrigeration and freezing units, or charging system for fleets of electric vehicles. The MPPT can be devoid of intermediate energy storage between the PV array and the plurality of loads.

[0011] In another aspect, embodiments relate to a multi-converter maximum power point tracking system comprising a photovoltaic (PV) array configured to generate electrical power. A controller is in communication with the PV array7and configured to execute a maximum power point tracking (MPPT) algorithm. A plurality' of variable direct current / direct current (DC / DC) converters are downstream of the controller, with each variable DC / DC converter configured to receive a weighted duty cycle signal from the controller. A plurality of loads are provided, with each load powered by a respective variable DC / DC converter. A feedback control system is configured to adjust weighting factors for the duty' cycle signals based on individual load requirements, wherein the controller is configured to coordinate power distribution among the plurality of loads while maintaining maximum power extraction from the PV array.

[0012] One or more of the following features can be included. The maximum power point tracking algorithm can comprise a perturb and observe algorithm. The plurality of loads can comprise an electrodialysis stack and a pump system. The electrodialysis stack can comprise alternating cation exchange membranes and anion exchange membranes arranged between diluate and concentrate flow channels. The feedback control system can be configured to maintain operation of the electrodialysis stack at or below a limiting current threshold based on measured flow rate and conductivity parameters.

[0013] In another aspect, embodiments relate to a method of driving a plurality of loads comprising drawing power from a photovoltaic (PV) array. The method includes using a maximum power point tracker (MPPT) in communication with the PV array and a plurality'Attorney Docket No.: MIT 25641 USPCT | 88212-431989 of power converters disposed downstream of the MPPT to regulate a power delivered to one or more loads located downstream of the plurality of power converters.

[0014] One or more of the following features can be included. The method can further comprise optimizing performance of the one or more loads using the plurality of power converters. Optimizing performance can further comprise optimizing a power delivered to each of the plurality of loads. The method can further comprise tracking the power delivered to each of the plurality of loads and a maximum power point of the PV array simultaneously. The plurality of loads can comprise an electrodialysis (ED) stack. The plurality of loads can be non-linearly related. A first load of the plurality of loads can be a hydraulic load and a second load of the plurality of loads can be an applied electrical load. The method can further comprise using a feedback loop to adjust a weighting factor on a duty cycle on a power converter of the plurality of power converters. The power drawn from the PV array can be not buffered by a battery. The power from the PV array can not pass through a converter upstream of the MPPT.

[0015] In another aspect, embodiments relate to a method of operating multiple loads from a photovoltaic source comprising drawing power from a photovoltaic (PV) array using a maximum power point tracker (MPPT). The method includes generating a common duty cycle command based on maximum power point tracking of the PV array. Individual weighting factors are applied to the common duty cycle command to create weighted duty cycle signals for a plurality of power converters. The weighted duty cycle signals are delivered to the plurality of power converters to independently regulate power to a plurality of loads. The weighting factors are adjusted based on feedback from the plurality' of loads to optimize performance of each load while simultaneously tracking the maximum power point of the PV array.

[0016] One or more of the following features can be included. The plurality of loads can comprise an electrodialysis stack and a pump system. Adjusting the weighting factors can comprise maintaining operation of a load of the plurality of loads at or below a limiting current threshold based on measured flow rate and conductivity parameters. The limiting current threshold can be calculated using an empirical model that relates limiting current to solution conductivity’ and flow rate.Attorney Docket No.: MIT 25641 USPCT | 88212-431989BRIEF DESCRIPTION OF DRAWINGS

[0017] This disclosure will be more fully understood from the following detailed description taken in conjunction with the accompanying drawings, in which:

[0018] FIG. 1A is a schematic diagram of a canonical, prior art off-grid system configuration that includes a maximum power point tracker (MPPT);

[0019] FIG. IB is a schematic diagram of a multi-converter MPPT architecture of the present embodiments;

[0020] FIG. 2 is a block diagram of a generalized direct-drive multi-converter MPPT control architecture of the present embodiments;

[0021] FIG. 3 is a cross-sectional view of an electrodialysis stack that can be used with the direct-drive multi-converter MPPT control architecture of FIG. 2;

[0022] FIG. 4 is a block diagram of an off-grid direct-drive control configuration using a multi-converter maximum power point tracker;

[0023] FIG. 5A is a perspective view of an experimental system setup for a laboratory-scale electrodialysis desalination system;

[0024] FIG. 5B is a magnified top view of a printed circuit board (PCB) used with the experimental system setup of FIG. 5 A;

[0025] FIG. 5C is a detailed perspective view of the electrodialysis desalination system of FIG. 5A;

[0026] FIG. 6A is a graph showing voltage tracking performance over time for the multiconverter MPPT system operating with an electrodialysis desalination setup;

[0027] FIG. 6B is a graph showing current tracking behavior over time for the multiconverter MPPT system operating with an electrodialysis desalination setup;

[0028] FIG. 6C is a graph showing flow rate stability over time for the multi-converter MPPT system operating with an electrodialysis desalination setup;

[0029] FIG. 6D is a graph showing conductivity measurements over time for the multiconverter MPPT system operating with an electrodialysis desalination setup;

[0030] FIG. 6E is a graph showing stack current control performance over time for the multi-converter MPPT system operating with an electrodialysis desalination setup;Attorney Docket No.: MIT 25641 USPCT | 88212-431989

[0031] FIG. 6F is a graph showing output load voltages and duty cycles over time for the multi-converter MPPT system operating with an electrodialysis desalination setup; and

[0032] FIG. 7 is a block diagram of a computing device that can be used to implement various aspects of the systems and methods described herein.DETAILED DESCRIPTION

[0033] Certain embodiments will now be described to provide an overall understanding of the principles of the structure, function, manufacture, and use of the systems, devices, related components (e.g., photovoltaic arrays, maximum power point trackers, DC / DC converters, electrodialysis stacks, pumps, microcontrollers, sensors, and feedback control systems), and methods disclosed herein. One or more examples of these embodiments are illustrated in the accompanying drawings. Those skilled in the art will understand that the devices and methods specifically described herein and illustrated in the accompanying drawings are nonlimiting exemplary' embodiments and that the scope of the present disclosure is defined solely by the claims. The features illustrated or described in connection with one exemplary embodiment may’ be combined with the features of other embodiments. Such modifications and variations are intended to be included within the scope of the present disclosure. Further, to the extent features, components, converters, loads, or the like are described as being "first," "second," "third," etc., and / or "upstream," "downstream," "input," "output," etc., such numerical and / or location ordering / identifi cation is generally arbitrary, and thus such numbering can be interchangeable unless indicated or otherwise understood by those skilled in the art to not be interchangeable.

[0034] The figures provided herein are not necessarily to scale, although a person skilled in the art will recognize instances where the figures are to scale and / or what a typical size is when the drawings are not to scale. Further, to the extent that linear or circular dimensions or shapes are used or described herein, such dimensions are not intended to limit the types of shapes or sizes of such devices, components, etc. A person skilled in the art will recognize that an equivalent to such linear and / or circular dimensions or shapes can be easily’ determined for any geometric shape (e.g., references to widths and diameters being easily adaptable for circular and linear dimensions, respectively, by a person skilled in the art). While in some embodiments movement of one component is described with respect to another, a person skilled in the art will recognize that other movements are possible. Further, to the extent arrows are used to describe a direction a component can expand or move, theseAttorney Docket No.: MIT 25641 USPCT | 88212-431989 arrows are illustrative and in no way limit the direction the respective component can expand or move. A person skilled in the art will recognize other ways and directions for creating the desired tension or movement.

[0035] Still further, in the present disclosure, like-numbered components of various embodiments generally have similar features when those components are of a similar nature and / or serve a similar purpose, unless otherwise noted or otherwise understood by a person skilled in the art. To the extent the present disclosure includes prototypes, mock-ups, bench models, or the like, a person skilled in the art will recognize how to rely upon the present disclosure to integrate the techniques, systems, devices, and methods into a product, such as direct-drive photovoltaic desalination systems, multi-converter power electronics systems, and renewable energy7management systems. A number of terms may be used throughout the disclosure interchangeably but will be understood by a person skilled in the art. By way of non-hmiting example, the terms "power converter," "DC / DC converter," and "buck converter," can be used interchangeably with one another. Further, "maximum power point tracker," and "MPPT," can be used interchangeably with one another, as can "electrodialysis stack," and "ED stack." Moreover, "duty cycle," "duty command," and "weighting factor" may also be used interchangeably with one another. Further still, it will be appreciated that although features may be discussed with respect to one embodiment within the present disclosure, these features can be applied to every embodiment of the present disclosure where such feature would be supported.

[0036] To the extent terms like "approximately," "about," and "substantially" are used herein, a person skilled in the art will appreciate the scope those words convey in the context of their usage. In the context of multi-converter maximum pow er point tracking systems, obtaining a certain degree of voltage regulation, current control, power distribution, and / or duty cycle precision, among other electrical and control parameters may be difficult, and thus use of terms like "approximately," "about," and "substantially" is intended to address this difficulty. A person skilled in the art will understand what constitutes how close a particular dimension or placement should be to still fall within the spirit of the quantification and description provided for herein. Even in instances where such terminology is not used, and a dimension or electrical parameter just includes the number or term (e.g., "maximum" is used instead of "substantially maximum"), a person skilled in the art will appreciate that, unless explicitly indicated otherwise, terms like "approximately." "about," and "substantially" are applicable to those dimensions and terms as w ell. The foregoing notwithstanding, a personAttorney Docket No.: MIT 25641 USPCT | 88212-431989 skilled in the art will appreciate that terms like "approximately," "about," and "substantially" at least encompass voltages, currents, power levels, duty cycles, etc. that are ±10%, 10°, etc. of the provided amount, or encompass dimensions that are ±5%, 5°, etc. of the provided amount, unless indicated otherwise or otherwise known to those skilled in the art. The present disclosure appreciates that a person skilled in the art, in view of the present disclosure, understands suitable placements for various features of the disclosed systems, devices, etc., and related components of any of the same, and thus to the extent a particular electrical parameter is described, unless it is explicitly indicated that electrical parameter, a person skilled in the art will appreciate other electrical parameters that are possible without impacting the overall system, device etc.

[0037] The present disclosure relates to system architectures of integrated multi-converter, direct-drive maximum power point trackers for operating loads from photovoltaic arrays to provide power to multiple loads downstream. In particular, the present disclosure generally provides for a system that includes an algorithm and physical prototype architecture that facilitates photovoltaic electrodialysis desalination without energy storage. The algorithm can accomplish at least two goals: (1) extracting the maximum power from the solar panels; and (2) producing the maximum amount of water by operating with optimal distributions of power to the hydraulic and electric loads on the electrodialysis desalination system. The system can include an integrated maximum power point tracker (MPPT) and optimal electrodialysis desalination controller all-in-one to facilitate operation without energy storage. The system of the present embodiments can use various control strategies and can be generalized to broader multi-load, variable power source applications that desire no energy storage. Challenges of using two or more converters (multiple loads), as opposed to a single converter like in conventional systems, in the system of the present embodiments, coupled with the maximum power point tracker, can include balancing variable power for multiple, individual loads (without energy storage), which are discussed and rectified in greater detail below.

[0038] A person skilled in the art wall recognize that maximum power point trackers (MPPTs) are algorithms that can pull the maximum power from a solar array despite intermittencies and variations due, at least in part, to the aforementioned factors altering the characteristic IV curves. MPPTs function by algorithmically determining the optimal voltage and current combination to be delivered as a duty to converters with which the MPPTs are in communication. For example, MPPTs are often coupled with a power converter, which canAttorney Docket No.: MIT 25641 USPCT | 88212-431989 convert the photovoltaic panel voltage to an applied load, often a battery in off-grid scenarios. The output voltage of this converter to the battery can be relatively constant and steady, while the input voltage from the photovoltaic panel can vary depending, for example, on the current drawn from the panel. This current can vary based on, for example, the duty cycle to switches in the converter, which can be, for example, metal oxide field effect transistors (MOSFETs) and / or insulated-gate bipolar transistors (IGBTs). A wide gamut of MPPT algorithms exist, however two common and conceptually simple algorithms include fractional open circuit voltage (FOCV) and perturb & observe (P&O). FOCV is a conventional, in-direct method that measures only the panel voltage and aims to keep the panel voltage at a fixed fraction of the known, open circuit voltage. This method, while being quite simple and highly robust, fails to track the maximum power point — especially with changes in temperature and / or shading. FOCV can be implemented with a simple feedback loop with reference on the target fractional panel voltage and process variable on the measured panel voltage. P&O is another common technique that is a hill-climbing method that iteratively compares the power to the previous timestep to see if there is an increase or decrease, and then compares the voltage to the previous timestep to determine the direction in which the algorithm had been traveling. Generally, MPPT algorithms can measure voltage, current, solar irradiance, and the panel array temperature, as well as in more complicated cases make weather predictions, and discretize these measurements into pieces of a broader array.

[0039] FIG. 1A illustrates an example embodiment of a canonical, prior art off-grid system configuration 10 that demonstrate how typical maximum pow er point trackers (MPPTs) operate in conventional photovoltaic systems. As shown, the system 10 can include a PV array 12, an MPPT 14, a DC / DC converter 16, and a battery 18. Voltage and current can pass from the PV array 12 to the MPPT 14, which outputs a duty command to the DC / DC converter 16 that flows as power to the battery' 18. There is no communication between the MPPT 14 and the one or more fixed DC / DC converters 20. Rather, the battery 18 can be in communication with the one or more fixed DC / DC converters 20, with downstream loads 22 operating therefrom with relatively fixed input and output voltages sourced by the constant battery' voltage. The loads 22 can be an ED system, though it can be a number of alternative systems, both in water purification / desalination. as well as powering of appliances and other technologies.Attorney Docket No.: MIT 25641 USPCT | 88212-431989

[0040] In contrast to the canonical configuration, the present disclosure provides a direct- drive system architecture 100 that eliminates the need for both a centralized DC / DC converter and battery storage. At least one novel feature of a system 100 of the present embodiments, illustrated in FIG. IB, includes creation and use of a controller that is in communication with a PV array 102, managing power distribution to dow nstream components through a dual buck converter configuration. The controller may coordinate the operation of multiple power conversion stages while executing both maximum power point tracking and optimal desalination control algorithms. In some embodiments, the controller can include a multiple load, direct-drive MPPT 104 for controlling a plurality of loads as a single unit. As noted above, the multiple load, direct-drive MPPT 104 can eliminate at least one converter 16, which can increase the energy efficiency and decrease the overall cost of the system. An additional benefit of the direct-drive operation of the system 100 can be the ability to eliminate the need for the battery 18. Batteries can increase the capital and operating cost of renew able systems, and can have detrimental beginning and end-of-life environmental effects.

[0041] As shown in FIG. IB, the proposed multi-converter MPPT architecture 100 can enable multiple loads to be directly-driven by a photovoltaic (PV) array 102 to the MPPT 104 to monitor and control power extraction from the photovoltaic source, rather than buffered by a battery or other energy7storage. The voltage and current can pass from the PV array 102 to the MPPT 104, which outputs a duty command to one or more variable DC / DC converters 110 that are in communication with downstream loads 112 operating therefrom. That is. the duty commands to the DC / DC converters 110 can be sourced directly from a supervisory7controller, e.g., the multi-converter MPPT 104, and the power sources for these converters 110 can come directly from the PV array 102, allowing variable control of these converters 110. Combining control schemes using this architecture can allow for control of variable loads using the variable DC / DC converters 110 to ramp up and / or ramp down their power.

[0042] The MPPT 104 may be configured to regulate both the downstream plurality of loads and a power point of the PV array 102 simultaneously. This dual regulation capability can allow the system 100 to extract maximum power from the photovoltaic source 102 while optimally distributing that pow er among multiple loads according to their individual operational parameters. Moreover, the proposed system architecture 100 eliminates the traditional centralized approach by enabling direct-drive operation where duty commands to the DC / DC converters are sourced directly from a supervisory controller and power sourcesAttorney Docket No.: MIT 25641 USPCT | 88212-431989 for these converters come directly from the photovoltaic array 102. In some embodiments, no converter exists upstream of the MPPT 104, which distinguishes the system from conventional configurations that include upstream power conditioning stages. The PV array 102 and the MPPT 104 may not be in communication with a battery, thereby eliminating the energy storage component that ty pically sen es as an intermediate buffer in traditional off- grid photovoltaic systems.

[0043] Referring to FIG. 2, a generalized direct-drive multi-converter MPPT control architecture 200 illustrates how multiple power converters can receive coordinated control signals while maintaining independent load regulation capabilities. The system architecture 200 can include a photovoltaic array 202 that provides input power to an MPPT system 104, which incorporates a generalizable MPPT algorithm 204 that processes photovoltaic array variables such as current, voltage, and temperature. The MPPT 104 can generate a feedback control signal that can be distributed to multiple downstream converters 220, enabling simultaneous maximum power point tracking and individual load optimization.

[0044] The generalized multi-converter MPPT architecture 204 can include a plurality of process setpoints 212, w ith each process setpoint being associated with a respective feedback controller 210. A first process setpoint 212a may feed into a first feedback controller 210a, which can receive sensed process data and generate a control signal based on the specific operational parameters of the associated load. The feedback controller output can be combined with a weighting factor, which may be represented as cu, and this weighting factor can be multiplied with a duty cycle signal Di to create a w eighted control signal, as discussed in greater detail below. This w eighted signal can control a first DC / DC converter 220a that supplies power to a first load 222a with the first DC / DC converter 220a and first load 222a being connected in a feedback loop where additional sensing provides information back to the first process setpoint 210a.

[0045] The architecture 200 can be extended to accommodate multiple loads through parallel control paths. A second process may include a second process setpoint 212b, a second feedback controller 210b receiving sensed process data, a weighting factor a.2. and a duty cycle signal D2 that controls a second DC / DC converter 220b supplying a second load 222b. This configuration can be repeated for additional processes, with the system supporting an Nth process that includes an Nth process setpoint 212n, Nth feedback controller 21 On, weighting factor an, duty cycle signal Dn, Nth DC / DC converter 220n, and Nth load 220n. Each converter-load pair can operate with its own feedback loop, whereAttorney Docket No.: MIT 25641 USPCT | 88212-431989 additional sensing from each load provides information back to its respective process setpoint.

[0046] The system architecture 200 can include a feedback loop that includes a weighting factor on a duty cycle on at least one of the plurality of power converters. The weighting factors can enable independent power allocation adjustments based on individual load objectives while maintaining coordination through the centralized MPPT 104. For example, if a load 222 such as a refrigeration unit begins to deviate from its target temperature setpoint 212, the feedback controller 210 may increase the associated weighting factor to allocate a greater portion of total extracted power to that specific load. The weighting factors can be dynamically adjusted in real-time based on sensed process variables and load-specific control objectives.

[0047] The power from the photovoltaic array 202 can flow directly to each DC / DC converter 220, with each converter independently regulating power delivery' to its associated load 222 based on the weighted duty cycle signals derived from both the common MPPT output and individual feedback control loops. This architecture can allow for independent control of multiple loads while maintaining coordination through the centralized MPPT system 104. The common command signal from the generalizable MPPT algorithm 204 can provide a baseline duty cycle that reflects the optimal operating point for maximum power extraction from the photovoltaic array 202.

[0048] The generalizable MPPT algorithm 204 can be implemented using various control methods to optimize power extraction from the photovoltaic source 202. The generalizable MPPT algorithm 204 can be implemented using fractional open circuit voltage (FOCV) methods, which measure panel voltage and maintain the panel voltage at a fixed fraction of the known open circuit voltage. FOCV methods can provide robust operation with simple implementation parameters, making them suitable for applications where computational resources may be limited. Alternatively, the generalizable MPPT algorithm 204 can be implemented using perturb and observe (P&O) methods, which employ hill-climbing techniques that iteratively compare power levels between successive time steps. P&O methods can determine the direction of voltage or current perturbations by analyzing whether power increases or decreases following each adjustment, allowing the algorithm to converge toward the maximum power point through systematic exploration of the operating space.Attorney Docket No.: MIT 25641 USPCT | 88212-431989

[0049] The generalizable MPPT algorithm 204 can incorporate measurements of voltage, current, solar irradiance, and panel array temperature to optimize power extraction under varying environmental conditions. In more advanced implementations, the algorithm 204 may incorporate weather prediction data and discretize measurements across broader array configurations to enhance tracking performance. The algorithm 204 can output reference voltages or duty cycle commands that sen e as the foundation for the weighted control signals distributed to each downstream converter. The combination of centralized maximum power point tracking with distributed load-specific control can enable optimal system performance across diverse operating conditions while maintaining the flexibility7to accommodate varying load parameters and priorities.

[0050] ELECTRODIALYSIS STACK

[0051] The system architecture 100 can include a plurality of loads that includes a hydraulic load and an applied electrical load. The hydraulic load may include pumping systems that circulate water through an electrodialysis stack 300 and associated piping, valves, and flow control components. The applied electrical load can include the electrodialysis stack itself, which consumes electrical power to drive the ion transport process through the application of voltage across the electrodes. The plurality of loads can include an electrodialysis (ED) stack or a pump, with these components representing the primary power-consuming elements in a direct-drive desalination system. The pump can provide the hydraulic energy7necessary to maintain fluid flow through the membrane stack, while the ED stack 300 can consume electrical energy7to create the driving force for ion separation and water desalination.

[0052] Referring to FIG. 3, an electrodialysis (ED) stack 300 structure illustrates the fundamental components and operating principles of electrochemical desalination systems that can serve as loads in the multi-converter MPPT architecture 104. FIG. 3 shows an electrodialysis stack 300 in cross-section, revealing the internal arrangement of membranes and flow channels that enable selective ion separation. In particular, the ED stack 300 can include a diluate (diluted, or less concentrated) stream 302 and a concentrate feed 304 of inlet and outlet solution. The diluate stream 302 may represent a diluted or less concentrated solution path where ions are removed from the feed water, while the concentrate stream 304 can carry a more concentrated solution where removed ions accumulate.

[0053] The stack configuration 300 can enable the separation of ionic species from feed solutions through the application of electrical potential across electrodes positioned atAttorney Docket No.: MIT 25641 USPCT | 88212-431989 opposite ends of the membrane assembly. For example, as shown, the ions within the stack 300 can pass through selectively permeable membranes based on charge, denoted as cation exchange membranes (CEM)s 306 and anion exchange membranes (AEMs) 308. The electrodialysis stack 300 can be unique in that the load can be a variable resistor, which varies with the solution concentration. In desalination, this resistance in particular can decrease with higher salt content, and contrastingly increases with lower salt content. That is, as the system desalinates closer to drinking water levels of salinity, the resistance can increase. A voltage can be applied to the electrodes from the stack (converter) power supply, driving current and the transport of salt ions through the membranes 306, 308, thus, desalinating the water. The diluate stream 302 can exit as a diluate output 312 with reduced ion concentration, producing desalinated water suitable for various applications. The concentrate stream 304 may exit as a concentrate output 314 with increased ion concentration creating a brine stream that contains the removed salts and other ionic species.

[0054] In some embodiments, the CEMs 306 may be positioned to allow passage of positively charged ions while blocking negatively charged ions, creating selective permeability based on ionic charge. The AEMs 308 can permit passage of negatively charged ions while preventing the transport of positively charged ions. The arrangement of alternating CEMs 306 and AEMs 308 between the diluate and concentrate channels can enable the separation and concentration of ions as solution flows through the stack. The selective permeability7of these membranes can be based on the electrical charge of the ionic species, with each membrane type allowing only ions of specific polarity to pass through.

[0055] An anode 316 can be positioned at one end of the electrodialysis stack 300, while a cathode 318 may be positioned at the opposite end to create the electrical field necessary for ion transport. The application of voltage between the anode 316 and the cathode 318 can drive current through the electrodialysis stack 300, creating an electrical potential that motivates the movement of ions through the selectively permeable membranes. The voltage application can drive current and salt ion transport for desalination by creating an electrical driving force that overcomes the natural resistance to ion movement through the membrane materials. The transport of salt ions through the selectively permeable ion-exchange membranes can result in the desalination of water as ionic species are selectively removed from the diluate stream 302 and concentrated in the concentrate stream 304.

[0056] The electrodialysis stack 300 can function as a variable resistor load that changes electrical characteristics based on solution concentration and operating conditions. InAttorney Docket No.: MIT 25641 USPCT | 88212-431989 desalination applications, the electrical resistance of the stack 300 may decrease with higher salt content in the feed solution, as increased ionic concentration provides more charge carriers to facilitate current flow. Conversely, the resistance can increase with lower salt content as the solution approaches drinking water levels of salinity, reducing the availability of charge carriers and increasing the electrical impedance of the system. This variable resistor nature can create dynamic loading conditions that affect power consumption and system performance throughout the desalination process.

[0057] The electrodialysis stack 300 operation can be constrained by a limiting current threshold that represents the maximum current density that can be applied without causing water dissociation. The limiting current constraint can occur when the boundary layer adjacent to the membrane becomes completely depleted of solute, creating conditions where excess voltage begins to cause dissociation of water molecules rather than productive ion transport. Operation beyond the limiting current can result in the generation of hydrogen and chlorine gas, which can reduce system efficiency and create safety concerns. The limiting current can vary nonlinearly with both the concentration of the solution in the diluate compartment and the water flow velocity, which uses dynamic control strategies to maintain optimal operation while avoiding overlimiting conditions.

[0058] There is significant interest in utilizing solar energy with electrodialysis in direct- dnve (battery -less) operation. Desalination without energy storage in general is a field of growing interest, as it reduces capital expenditure, volume, and operational complexity'. A significant challenge to direct-drive, PV electrodialysis, however, is optimally varying the two power supplies to maximize water production rate, while also tracking the maximum power point.

[0059] Therefore, a novel feature of the present embodiments is the creation of a two- converter, direct-drive maximum power point tracker that accomplishes both: (1) the optimal balancing between the two converter loads; and (2) tracking of the maximum power point, simultaneously. Specifically, the MPPT 104 is an integrated MPPT solar pump inverter with an electrodialysis power supply for the application of direct drive electrodialysis desalination. This allows the MPPT 104 to function while fully eliminating battery storage and eliminate a power converter in this system, while tracking maximum power and optimally balancing the two converters to produce the most water.Attorney Docket No.: MIT 25641 USPCT | 88212-431989

[0060] FIG. 4 illustrates a system architecture 400 of the present embodiments that is applied to the ED system 300 to form an off-grid direct-drive ED control configuration that uses a multi-converter MPPT 404. While typical MPPTs use one power converter to draw maximum power from a solar array to charge a battery which then supplies power to multiple loads dow nstream, the present architecture 400 can use tw o or more loads directly-driven by the photovoltaic array 402 rather than buffered by a battery or other energy storage. The integrated system 400 can eliminate the traditional energy storage component while maintaining optimal powder extraction and load regulation capabilities through coordinated control of multiple dow nstream converters. It will be appreciated that while the system architecture 400 is discussed herein as being applied to the ED stack load 300, the ED stack load 300 can be a load of either of the system architectures 100, 200 discussed above.

[0061] The integrated dual buck converter w ith MPPT 404 can include a plurality of buck converters that simultaneously optimize load balancing for maximum w ater production within the ED stack 300 while tracking the maximum power point of the photovoltaic source. The dual converter architecture may enable independent pow er regulation for different system components while maintaining overall system coordination through the centralized MPPT control algorithm. The plurality7of buck converters can be configured to operate with variable duty cycles that respond to both maximum power point tracking parameters and individual load optimization objectives.

[0062] As shown in FIG. 4. a generalized P&O algorithm 405 of the MPPT 404 can be used to read the current and voltage from the photovoltaic array 402, outputting a reference voltage. This operating (reference) voltage can be tracked using feedback control by measuring the PV array voltage and altering an overarching (common command) output duty7cycle signal, DI. DI can be fed directly to the pump power converter 407 to directly increase or decrease the power consumption and flow rates. DI can be simultaneously biased by a weighting factor, a, which can alter the duty7to the electrodialysis stack 300 powder converter, a can be adjusted, for example at least based on a feedback loop, or ED limiting current model 420, which aspires to command an electrodialysis stack current, istacfc which matches the limiting current. iiim- The limiting current can be the point at which water dissociates, and can be determined by inputting measured conductivity, C, and flow rate, Q, into an empirical model. This can be the individual, subobjective for the electrodialysis stack converter (match limiting and commanded current, to maximize desalination rates), while theAttorney Docket No.: MIT 25641 USPCT | 88212-431989 individual, sub-objective of the pump is to maximize the flow rate at any given time. Both sub-objectives can be used for optimal desalination rate. Both loads, e.g, the pump 407 and the ED stack 300, can concurrently consume power, which then affects the solar array voltage and current, and thus is fed back into the MPPT 404. This closed loop control scheme allows for both maximum power point tracking and continuing to operate at the optimal desalination rate.

[0063] The nested feedback loop can use a continuously adjusted weight on one converter's duty cycle, with the rest of the control being algorithmically similar to canonical voltage setpoint tracking and perturb and observe methods. A proportional-integral-derivative (PID) controller 422 may process the difference between the commanded limiting current and the actual stack current to adjust the weighting factor a through dynamic feedback control. The weighting factor can modify the duly cycle signal sent to the ED stack converter 424, allowing independent control of the stack current while the system tracks the maximum power point of the photovoltaic array. The nested control architecture can enable simultaneous optimization of both maximum power extraction and optimal desalination performance without using intermediate energy storage.

[0064] The system 400 can include multiple pumps such as the main pump 407 for bulk fluid flow and an electrode rinse pump 409 for clearing gaseous formations from faradaic reactions. The main pump 407 may circulate feed water through the diluate and concentrate channels of the electrodialysis stack 300, providing the hydraulic energy necessary to maintain proper flow rates for optimal ion transport and desalination performance. The electrode rinse pump 409 can provide a separate flow stream to the electrode compartments, helping to remove hydrogen and oxygen gases that may form during the electrodialysis process and could otherwise interfere with system operation. The electrode rinse pump 409 may prevent gas accumulation at the electrodes that could reduce system efficiency or create operational instabilities.

[0065] Each load associated with each of the plurality of power converters may be configured to be operated simultaneously within the system architecture 400. The simultaneous operation capability can enable coordinated power management where the pump loads 407 and the electrodialysis stack load 300 operate concurrently while drawing power from the same photovoltaic source. The plurality of power converters can regulate power delivery to their respective loads in real-time, with each converter responding to both the common MPPT control signal and individual load-specific feedback parameters. TheAttorney Docket No.: MIT 25641 USPCT | 88212-431989 simultaneous operation of multiple loads can be facilitated by the independent duty cycle control for each converter, allowing the system to balance power allocation among different loads while maintaining maximum power point tracking performance.

[0066] The plurality of power converters can be variable direct current (DC) / DC converters that provide adjustable output voltage and current based on the duty cycle commands received from the MPPT controller. The variable DC / DC converter capability can enable dynamic power allocation among multiple loads while maintaining optimal power extraction from the photovoltaic source. Each variable DC / DC converter may respond independently to its respective duty cycle command, allowing for simultaneous optimization of multiple load parameters within the same system. The variable nature of these converters can distinguish them from fixed-ratio converters that provide constant voltage or current outputs regardless of input conditions or control commands.

[0067] Both loads can draw power concurrently from the photovoltaic array 402, with their combined power consumption affecting the array voltage and current in a manner that feeds back into the MPPT algorithm 404. The integrated control system 400 may coordinate the operation of multiple converters to ensure that the total power consumption matches the available power from the photovoltaic source while optimizing the distribution of that power among the various system loads. The simultaneous operation of the pump 407 and electrodialysis stack loads 300 can enable continuous desalination operation that adapts to varying solar conditions without energy storage to buffer power fluctuations.

[0068] The power converter hardware implementation can utilize various converter topologies to accommodate different application parameters and operating conditions. The MPPT sy stem may be implemented with buck-boost, flyback, and fly-buck converter topologies in addition to the dual buck converter configuration. Buck-boost converters can provide the capability to either step up or step down the input voltage depending on the load parameters and photovoltaic array operating conditions. Flyback converters may offer galvanic isolation between the input and output, which can be advantageous for certain safety-critical applications. Fly-buck converters can combine the benefits of both fly back and buck topologies, providing multiple isolated outputs from a single converter stage.

[0069] The system 400 may utilize galvanically isolated DC / DC converters for safety in industrial-scale, high voltage electrodialysis applications where water leakage poses operator shock risk. Galvanic isolation can provide electrical separation between the photovoltaicAttorney Docket No.: MIT 25641 USPCT | 88212-431989 input and the load outputs, preventing current flow through unintended paths that could create safety hazards. In industrial electrodialysis systems that operate at hundreds of volts, the presence of conductive water solutions can create pathways for electrical current that could endanger operators. Isolated converter topologies can eliminate these safety concerns by ensuring that no direct electrical connection exists between the high-voltage photovoltaic source and the water-handling components of the system.

[0070] For home-scale applications operating at lower voltages, the dual buck converter architecture may be selected to provide a simpler and more cost-effective solution. The plurality of power converters can comprise a dual buck converter that provides independent power regulation for multiple loads while maintaining a straightforward control implementation. The dual buck converter configuration can eliminate the complexity associated with transformer-based isolated topologies while still providing the necessary power conversion and regulation capabilities for residential-scale electrodialysis systems. The selection of the dual buck architecture can be based on the lower voltage parameters of home-scale systems, which ty pically operate at voltages that do not present the same safety concerns as industrial-scale installations.

[0071] The power converters may be implemented as synchronous buck converters using two N-channel MOSFETs with dead time control to prevent shoot-through. The synchronous buck converter topology can provide higher power efficiency compared to non-synchronous designs by replacing the freewheeling diode with an actively controlled MOSFET. The use of two N-channel MOSFETs can enable bidirectional current flow and reduce conduction losses during both the on-time and off-time portions of the switching cycle. Dead time control may be implemented to prevent simultaneous conduction of both MOSFETs. which could create a short circuit path from the input to ground and result in device failure or system damage.

[0072] The system can operate in continuous conduction mode (CCM) with specific current ripple constraints to ensure stable operation and minimize electromagnetic interference. In continuous conduction mode, the inductor current may never reach zero during the switching cycle, which can provide more predictable control characteristics and reduced current stress on the switching devices. The current ripple constraint can be expressed mathematically as zlt / 2 < (iL), where the peak-to-peak current ripple must be less than or equal to twice the average inductor current to maintain continuous conduction. This constraint can ensure that the minimum inductor current remains positive throughout the switching cycle, preventingAttorney Docket No.: MIT 25641 USPCT | 88212-431989 the converter from entering discontinuous conduction mode, though it will be appreciated that the converter can operate in discontinuous conduction mode in some embodiments.

[0073] The inductor sizing for the buck converters can be determined using the fundamental voltage-current relationship for inductors. The basic inductor equation can be expressed as V = L where the voltage across the inductor is proportional to the rate of change of current through the inductor. For buck converter applications, this relationship can be rearranged to determine the minimum inductance to maintain the desired current ripple characteristics. The inductor sizing equation can be written asV°’('1 P),r, where Vois the output voltage, D is the duty cycle, T is the switching period, and (i0) is the average output current.

[0074] The system can include specific component values such as 1.124 mH inductors, 2000 pF capacitors, and operates at 10 kHz switching frequency. The 1.124 mH inductor value may be selected to provide adequate current ripple control while maintaining reasonable component size and cost for the home-scale application. The 2000 pF capacitor value can provide sufficient output filtering to meet the voltage ripple specifications for both the pump and electrodialysis stack loads. The 10 kHz switching frequency can represent a compromise between switching losses, which increase with frequency, and component size parameters, which decrease with higher switching frequencies.

[0075] The capacitor sizing for the output filters can be determined based on the current ripple and voltage ripple parameters of the system. The capacitor sizing equation can be expressed as C >where C is the capacitance, Ai is the inductor cunent ripple, T is the switching period, and AV is the allowable output voltage ripple. This relationship can account for the fact that the capacitor must supply current to the load during the portion of the switching cycle when the inductor current is below the average load current. The capacitor value can be selected to ensure that the voltage ripple remains within acceptable limits for proper load operation.

[0076] The system can operate with voltage ranges of about 0 to about 60 VDC input and about 0 to about 24 VDC output, with current ranges of about 0 to about 5 A input and about 0 to about 2 A output. The input voltage range may accommodate various photovoltaic array configurations, including single panels and small series-connected arrays typical of residential installations. The output voltage range can be suitable for standard pump andAttorney Docket No.: MIT 25641 USPCT | 88212-431989 electrodialysis stack operating voltages commonly used in home-scale water treatment systems. The current ranges may be sized to handle the power parameters of typical residential desalination loads while providing adequate margin for system variations and transient conditions.

[0077] The system may operate with voltage ripple specifications of 100 mV for pumps and 6 mA peak-to-peak current ripple for the electrodialysis stack. The 100 mV voltage ripple specification for pumps can prevent flow rate fluctuations that could lead to instabilities in the cunent control and concentration output from the electrodialysis stack. The 6 mA peak- to-peak current ripple specification for the electrodialysis stack 300 can prevent exceeding the limiting current threshold that could cause water dissociation and the generation of hydrogen and chlorine gas. These ripple specifications can ensure stable operation while maintaining safety margins for both hydraulic and electrochemical system components.

[0078] The MPPT can comprise a plurality of power supplies, with each power supply being in communication with a power converter of the plurality of power converters. Each power supply may provide the necessary control voltages and currents for the operation of its associated power converter, including gate drive signals for the switching devices and bias voltages for control circuitry . The plurality of power supplies can enable independent operation of multiple converters while maintaining coordination through the centralized MPPT control algorithm. The communication between each power supply and its associated converter can facilitate real-time adjustment of duty cycles and other operating parameters based on both maximum power point tracking parameters and individual load optimization objectives.

[0079] The system 400 may include point of load (POL) power converters such as Traco Power TSR 1-48120WI for 12V and TSR 1-4850WI for 5V control circuitry power. The POL converters can provide regulated power for the control electronics, sensors, and communication interfaces that enable the MPPT and load regulation functions. The 12V POL converter may supply power for gate drivers, relay coils, and other higher-power control components, while the 5V POL converter can provide power for microcontrollers, sensor interfaces, and digital communication circuits. The use of dedicated POL converters can ensure stable operation of the control systems regardless of variations in the main power conversion stages or load conditions.Attorney Docket No.: MIT 25641 USPCT | 88212-431989

[0080] The system 400 may incorporate various sensing and monitoring components to enable precise control and feedback for both maximum power point tracking and load optimization functions. The sensing architecture can provide real-time measurements of electrical parameters, hydraulic conditions, and water quality characteristics that are used by the control algorithms to maintain optimal system performance. The sensor selection and placement can be designed to provide accurate measurements while maintaining costeffectiveness and reliability for residential-scale applications.

[0081] The system 400 can include voltage and current sensors such as Texas Instruments INA237AIDGST devices for electrical parameter measurements throughout the power conversion and distribution network. The INA237AIDGST sensors may provide high- accuracy voltage and current measurements with integrated analog-to-digital conversion capabilities that facilitate direct communication with microcontroller-based control systems. These sensors can be positioned at multiple locations within the system to monitor both input conditions from the photovoltaic array and output conditions at each load converter. The voltage and current sensing capability can enable the MPPT algorithm to calculate instantaneous power levels and track changes in the maximum power point as environmental conditions vary’.

[0082] The photovoltaic input sensing can utilize INA237AIDGST sensors to monitor the voltage and current delivered by the solar array to the MPPT system. These input measurements may provide the fundamental data for maximum power point tracking algorithms, including perturb and observe methods that rely on power calculations derived from voltage and current measurements. The input sensing can enable the MPPT controller to determine the optimal operating point for the photovoltaic array and adjust the duty cycles of downstream converters accordingly. The high resolution and accuracy of the INA237AIDGST sensors can facilitate precise power calculations that are necessary for effective maximum power point tracking under vary ing solar conditions.

[0083] The stack converter output current sensing may employ INA237AIDGST sensors to monitor the current delivered to the electrodialysis stack for limiting current control and feedback regulation. The output current measurements can provide real-time feedback to the control system regarding the actual current flowing through the electrodialysis stack, which can be compared to the calculated limiting current setpoint to prevent water dissociation and maintain optimal desalination performance. The current sensing at the stack converter outputAttorney Docket No.: MIT 25641 USPCT | 88212-431989 can enable the feedback control loop that adjusts the weighting factor a to maintain the stack current at or below the limiting current threshold while maximizing desalination throughput.

[0084] The controller may receive sensor data from the flow meters, conductivity probes, and electrical measurements to execute the integrated MPPT and optimal desalination control algorithms. For example, the system can include flow rate monitoring using Dijiang YF- S201 flow sensors to measure the volumetric flow rate of water through the electrodialysis system. The controller can output pulse-width modulated signals to each buck converter, independently adjusting the duty cycles to optimize power allocation between the pumps and the electrodialysis stack. The YF-S201 flow sensors may operate on a pulse-frequency principle where the sensor generates digital pulses at a rate proportional to the volumetric flow rate passing through the sensor. The flow7rate measurements can be determined byreading the output pulse frequency from the flow meter and multiplying by a calibrated conversion factor to obtain flow rate values in appropnate units such as milliliters per minute. The flow rate sensing can provide input data for the limiting current calculation model and enable monitoring of hydraulic system performance.

[0085] The flow rate sensors may be positioned to monitor the diluate output stream 312 from the electrodialysis stack 300, providing measurements of the desalinated water production rate. The flow rate measurements can serve multiple functions within the control system, including input to the empirical limiting current model and monitoring of pump performance and hydraulic system operation. The How rate data can be averaged over multiple readings to eliminate noise and provide stable input values for the limiting current calculations that determine the optimal current setpoint for the electrodialysis stack 300.

[0086] The system can incorporate conductivity7sensing using Keyestudio KS0429 conductivity sensors to measure the salinity and ionic concentration of water streams within the electrodialysis system. The KS0429 conductivity sensors may provide analog output signals that scale linearly with the conductivity of the measured solution, enabling direct interface with analog-to-digital conversion systems in the control electronics. The conductivity measurements can provide information about both the feed water salinity and the desalinated product water quality, enabling assessment of desalination performance and calculation of salt removal efficiency.

[0087] The conductivity sensors may be positioned to monitor both the feed water stream entering the electrodialysis stack 300 and the diluate output stream 312 exiting the stackAttorney Docket No.: MIT 25641 USPCT | 88212-431989300. The feed water conductivity measurements can provide baseline salinity' information that affects the electrical characteristics and limiting current behavior of the electrodialysis stack 300. The diluate output conductivity measurements can indicate the quality of the desalinated water and the effectiveness of the ion removal process. The conductivity7measurements can serve as input parameters for the empirical limiting current model that determines the optimal operating current for the electrodialysis stack 300.

[0088] In some embodiments, a storage tank (not shown) may collect both the diluate and concentrate output streams 312, 314 from the electrodialysis stack 300. allowing them to remix and be recirculated as feed water for continuous operation during testing. The remixing arrangement can maintain constant feed water concentration throughout experimental testing by combining the desalinated product water with the concentrated brine stream. This continuous mode operation may enable extended testing periods without external feed water supplies or waste water disposal. The storage tank configuration can facilitate evaluation of system performance under steady-state conditions while maintaining consistent input water characteristics.

[0089] The limiting current calculation may utilize an empirical model that relates the maximum allowable current to the measured flow rate and conductivity parameters. The empirical limiting current model can be expressed mathematically as= a • C • Q , where iiim represents the limiting cunent, C represents the solution conductivity, Q represents the flow rate, and a and P represent empirically determined fitting coefficients. This mathematical relationship can capture the nonlinear dependence of limiting current on both hydraulic and electrochemical parameters that affect the ion transport characteristics of the electrodialysis stack.

[0090] The empirical fitting coefficients for the limiting current model may be determined through experimental characterization of the specific electrodialysis stack configuration used in the system. For a PC Cell 20 cell pair stack, the fitting coefficients can be determined as A = le-5 and B = 0.4 based on current-voltage curve analysis and limiting current measurements across various operating conditions. The coefficient A = le-5 may represent a scaling factor that accounts for the physical dimensions, membrane properties, and electrode characteristics of the specific stack design. The coefficient B = 0.4 can represent the power law relationship between flow rate and limiting current that reflects the influence of mass transfer limitations and boundary layer effects on ion transport.Attorney Docket No.: MIT 25641 USPCT | 88212-431989

[0091] The empirical model coefficients may be specific to the membrane materials, spacer geometry, number of cell pairs, and water composition characteristics of the particular electrodialysis stack implementation. The PC Cell 20 cell pair stack configuration can include specific coefficient values that account for the stack's physical dimensions, membrane surface area, and flow channel geometry. The empirical coefficients can be determined through experimental measurement of current-voltage curves at various flow rates and conductivity levels, with the limiting current identified as the point where the current-voltage relationship transitions from linear ohmic behavior to current-limited behavior.

[0092] The limiting current model can provide real-time calculation of the maximum allowable current based on continuously measured flow rate and conductivity' parameters. The model output can serve as the setpoint for the feedback control loop that adjusts the weighting factor a to maintain the electrodialysis stack current at or below the limiting current threshold. The empirical model approach can accommodate the complex, nonlinear relationships between operating parameters and limiting current behavior without detailed physical modeling of the electrochemical and mass transfer processes within the electrodialysis stack.

[0093] The sensor data acquisition and processing may involve averaging multiple readings to reduce noise and improve measurement stability for control system applications. The flow rate measurements can be averaged over 100 readings to eliminate noise for the limiting current calculation, while voltage and current sensors can be internally averaged by the integrated circuits over 32 readings to provide stable measurement values. The conductivity sensor readings may be processed using linear scaling analog input conversion to provide conductivity values in appropriate units for the limiting current model calculations. The sensor data processing can ensure that the control algorithms receive stable, accurate input values that enable reliable maximum power point tracking and optimal desalination control.

[0094] The system may be designed for home-scale brackish water applications with waters less than 2000 mg / L, representing typical salinity levels found in groundwater sources in regions such as India and the American Southwest. The home-scale application context can involve compact system packaging suitable for wall-mounted or under-sink cabinet installations in residential settings. The electrodialysis stacks for this application may be approximately the size of a lunchbox, providing a compact form factor appropriate for point- of-use water treatment systems. The pumps utilized can be relatively small positive-Attorney Docket No.: MIT 25641 USPCT | 88212-431989 displacement pumps similar to those used in home aquariums and garden water features, maintaining compatibility with residential power and space constraints.

[0095] The system may operate with a 50% recovery ratio where the diluate and concentrate channels have equivalent flow rates, providing balanced hydraulic operation and consistent product water quality. The equivalent flow rate operation can ensure that the hydraulic resistance of both channels remains balanced, preventing preferential flow that could reduce desalination efficiency or cause uneven membrane utilization. The 50% recovery ratio may represent an optimal balance between product water yield and energy efficiency for brackish w ater applications, maximizing the volume of desalinated water produced while maintaining reasonable power consumption levels. The balanced flow operation can facilitate stable control of the limiting current and optimal power allocation between the hydraulic and electrical loads within the system.

[0096] The experimental validation of the multi-converter maximum power point tracker system can be demonstrated through systematic testing using controlled solar simulation equipment and step response methodology. The experimental approach may utilize a combined pow er supply and solar simulator to create repeatable test conditions that enable evaluation of both maximum power point tracking performance and optimal desalination control under varying photovoltaic array conditions. The testing methodology can provide quantitative validation of the system's ability to simultaneously regulate multiple loads while maintaining optimal power extraction from the photovoltaic source.

[0097] The control software and algorithms can be implemented digitally on a microcontroller to provide an integrated maximum power point tracker (MPPT) and optimal electrodialysis desalination controller that facilitates operation without energy storage. The microcontroller implementation may utilize a Teensy 4. 1 processor that provides sufficient computational capability and input / output resources to execute the control algorithms while interfacing with multiple sensors and powder converters. The digital implementation can enable precise timing control, real-time sensor data processing, and coordinated management of multiple power conversion stages within a single integrated control system.

[0098] The control system can be implemented on a microcontroller such as Teensy 4.1 running at approximately 2.7 ms per cycle with specific PID coefficients. The control loop timing may be determined by the maximum computational rate that the microcontroller can handle w hile executing sensor reading, algorithm processing, and output generationAttorney Docket No.: MIT 25641 USPCT | 88212-431989 functions. The 2.7 ms cycle time can provide sufficient bandwidth for tracking dynamic changes in solar conditions and load parameters while maintaining stable control performance. The control system may output sensor data over USB serial communication to a computer using a Python script every 10 cycles, enabling real-time monitoring and data logging of system performance parameters.

[0099] The control system can incorporate multiple protection mechanisms to ensure safe operation under various operating conditions and fault scenarios. Undervoltage protection may be implemented with a threshold of less than 17 VDC on the photovoltaic input, as the 12 VDC point-of-load converters that supply the gate drivers may not be rated to operate below this voltage level. The system can include duty cycle limits that prevent excessively high duty cycles greater than 90% that could saturate the buck converters and cause instability. Low voltage shutoff conditions may be implemented when the commanded duty cycle falls below 5%, indicating insufficient solar power availability for proper system operation.

[0100] The sensor reading methodology can utilize I2C communication protocols with internal averaging to provide stable and accurate measurements for the control algorithms. The current and voltage sensors may be read over I2C communication, with readings internally averaged by the integrated circuits over 32 samples to reduce noise and improve measurement precision. The output current and voltage sensors can be read every 8 control cycles to monitor load conditions and converter performance. The input current and voltage sensors may be read every’ other cycle to provide frequent updates for the MPPT algorithm while balancing computational load and communication bandwidth parameters.

[0101] The MPPT algorithm implementation can utilize voltage setpoint tracking with PID feedback control to maintain optimal power extraction from the photovoltaic array. The MPPT algorithm may be executed every 4000 loop checks, which can provide a stable time period between MPPT algorithm updates that allows the input voltage tracking PID loop to reach and settle at the MPPT setpoint. The algorithm timing can be conserv atively slow7to ensure stability while demonstrating that the control scheme combining MPPT operation with independent downstream converter objectives remains feasible. The MPPT algorithm can adjust the input voltage setpoint in increments of 0.5 V to provide sufficient change in overall power above the sensor noise floor for proper algorithm interpretation.Attorney Docket No.: MIT 25641 USPCT | 88212-431989

[0102] The voltage setpoint range for the MPPT algorithm may be constrained between 18 VDC minimum and 26 VDC maximum to ensure compatibility with downstream load parameters. The minimum voltage setpoint of 18 VDC can be established just above the minimum operating voltage for the point-of-load converters that supply control power. The maximum voltage setpoint of 26 VDC may be selected to ensure that the pump loads do not exceed their maximum rated voltage of 24 VDC when operating at maximum duty cycle of 90%. The voltage setpoint can be initialized at 18 VDC at system startup and adjusted by the MPPT algorithm based on power optimization calculations.

[0103] The PID feedback control for voltage setpoint tracking can utilize specific coefficients that provide stable control response while maintaining adequate tracking performance. The PID coefficients for the voltage tracking loop may be configured as Kp = 10, Ki = 1, and Kd = 0, providing proportional and integral control action without derivative control. The PID controller can operate with 8-bit resolution to alter the overall duty signal that serves as the foundation for power distribution to multiple downstream converters. The voltage tracking PID loop may measure the photovoltaic array voltage and compare it to the voltage setpoint established by the MPPT algorithm, adjusting the common duty cycle signal to maintain the desired operating point.

[0104] The PWM generation for the power converters can be implemented using digital control with precise timing characteristics to ensure proper switching operation. Square waves with differing duty cycles may be created and digitally driven using interrupt-based timing to ensure consistency at a switching frequency of 10 kHz. The PWM generation can implement a dead time fraction of 10% of the total switching period, corresponding to approximately 10 pS of dead time between high-side and low-side MOSFET switching transitions. The dead time implementation may prevent shoot-through conditions that could damage the sw itching devices by ensuring that both MOSFETs in each converter leg are never conducting simultaneously.

[0105] The step increment approach for the MPPT algorithm can utilize perturb and observe methodology that systematically explores the power-voltage characteristic of the photovoltaic array. The algorithm may compare the current power level to the previous measurement and determine whether the voltage perturbation resulted in increased or decreased power extraction. Based on this comparison, the algorithm can determine the direction for the next voltage perturbation to continue climbing toward the maximum power point. The stepAttorney Docket No.: MIT 25641 USPCT | 88212-431989 increment size of 0.5 V can provide sufficient resolution for effective maximum power point tracking while avoiding excessive oscillation around the optimal operating point.

[0106] The optimal desalination algorithm can read flow and conductivity sensors every control cycle to provide continuous monitoring of hydraulic and water quality parameters. The sensor readings may be averaged over 1 0 samples to eliminate measurement noise and provide stable input values for the limiting current calculation. The flow meter can be read using frequency measurement techniques where analog input signals are converted to digital pulses when the signal crosses predetermined threshold levels. The conductivity sensor may be read using linearly scaling analog input conversion that provides conductivity values proportional to the analog voltage output from the sensor.

[0107] The stack current control can be implemented using a dedicated PID loop that adjusts the weighting factor a to bias the duty cycle received from the MPPT algorithm. The stack current PID controller may utilize coefficients Kp = 100, Ki = 100, and Kd = 0, providing aggressive proportional and integral control action to maintain tight regulation of the electrodialysis stack current. The weighting factor a can be constrained within the range 0 < a < 2, which physically means that the electrodialysis stack duty cycle can be no more than two times the duty cycle commanded to the pump converter. The weighting factor constraint may prevent excessive power allocation to the stack that could compromise pump operation or overall system stability.

[0108] The integrated control architecture can enable simultaneous regulation of both the plurality of loads and the power point of the photovoltaic array through coordinated algorithm execution. The system architecture may accomplish simultaneous regulation by combining the MPPT voltage setpoint tracking with individual load optimization objectives within a single control framework. The plurality of loads and the power point of the PV array can be regulated simultaneously through the nested feedback control structure that maintains maximum power extraction while optimizing power distribution among multiple loads. The simultaneous regulation capability' can distinguish the system from conventional approaches that separate maximum power point tracking from load regulation functions.

[0109] The control strategies implemented in the electrodialysis desalination system can be generalized to broader multi-load, variable power source applications that desire operation without energy storage. The fundamental control architecture may be adapted to accommodate different load types, power levels, and optimization objectives whileAttorney Docket No.: MIT 25641 USPCT | 88212-431989 maintaining the core principle of coordinated maximum power point tracking with distributed load regulation. Applications such as solar-powered refrigeration systems, multi-pump chemical processing systems, and direct-drive electrolyzer installations can benefit from similar control approaches that eliminate energy storage parameters while optimizing power allocation among multiple loads.

[0110] The generalized control framework can accommodate various load characteristics and optimization objectives through modification of the weighting factors, PID coefficients, and constraint ranges. Multi-load systems with different power parameters may utilize similar nested feedback control structures where each load receives a weighted duty cycle signal derived from the common MPPT output. The weighting factors can be adjusted based on load-specific feedback signals and optimization objectives, enabling independent control of multiple loads while maintaining coordination through the centralized maximum power point tracking algorithm. The control strategies can provide a foundation for developing direct-drive renewable energy systems across diverse applications where energy storage elimination represents a significant advantage in terms of cost, complexity, and environmental impact.

[0111] Referring to FIGS. 5A-5C, an embodiment of an experimental system setup 500 can demonstrate the integration of power electronics and hydraulic components for testing the multi-converter maximum power point tracker with an electrodialysis desalination system 600 that includes the ED stack 300. The laboratory-scale setup 500 may include a printed circuit board (PCB) 502 positioned to interface with various system components and provide centralized control functionality. The PCB 502 can contain multiple electronic components, as shown in FIG. 5B, including a microcontroller 504 for executing control algorithms, low voltage regulators 506 for providing stable power to control circuits, a relay 508 for switching operations, input filtering components 510 for power conditioning, and multiple buck converters 512 labeled as Buck 1, Buck 2, Buck 3, and Buck 4 for independent load regulation. Adjacent to the PCB 502, a laptop 514 may be positioned for data acquisition and system monitoring, providing an interface for real-time observation of system performance parameters and algorithm execution.

[0112] The experimental setup 500 can include a PV simulator 516 implemented as a rackmounted unit that provides simulated solar panel input to the system for controlled testing conditions. The PV simulator 516 may enable precise control of photovoltaic array characteristics including voltage, current, and power output parameters that would otherwiseAttorney Docket No.: MIT 25641 USPCT | 88212-431989 vary' with environmental conditions in outdoor testing scenarios. The controlled simulation environment can facilitate systematic evaluation of the multi-converter MPPT performance across defined operating ranges and step response conditions. The PV simulator 516 can provide repeatable test conditions that enable validation of the control algorithms and power distribution strategies under various simulated solar irradiance levels.

[0113] The central portion of the experimental setup 500 may feature an electrodialysis system mounted 600 within a transparent enclosure that provides visibility of the internal components while containing the hydraulic system. The electrodialysis system 600, which can include the ED stack 300 and other components, can include pumps 602 and valves 604 for fluid management, with these components arranged to provide controlled water circulation through the membrane stack. A beaker 606 may be positioned below the electrodialysis system 600 for water collection and recirculation, serving as a feed water storage reservoir that maintains continuous operation during testing. The transparent enclosure can enable visual monitoring of fluid flow patterns, bubble formation, and other operational characteristics that may affect system performance.

[0114] The electrodialysis stack 300 may be visible within the enclosure along with associated hydraulic connections that facilitate water flow through the diluate and concentrate channels. The hydraulic connections can include inlet and outlet fittings that direct feed water into the stack and collect the separated product and waste streams. The stack positioning within the enclosure may provide access for electrical connections to the electrodes while maintaining proper hydraulic sealing to prevent leakage. The integration of electrical and hydraulic connections can enable simultaneous monitoring of both power consumption and water processing performance during experimental operation.

[0115] With continued reference to FIGS. 5A-5C, the electrodialysis system 600 can include sediment filters 603 that may be mounted vertically within the system 600 to provide pretreatment of the feed water before it enters the electrodialysis stack. The sediment filters can remove particulate matter and suspended solids that could otherwise interfere with membrane performance or cause fouling of the ion exchange surfaces. The vertical mounting arrangement may facilitate easy replacement of filter elements during maintenance operations while minimizing the footprint of the filtration system.

[0116] FIG. 5C illustrates the ED system 600 in greater detail. As shown, the system 600 can include a main pump 607 that provides bulk fluid circulation through the diluate andAttorney Docket No.: MIT 25641 USPCT | 88212-431989 concentrate channels of the electrodialysis stack 300. The main pump 607 may be sized to deliver the flow rates necessary for optimal ion transport and desalination performance while operating within the power constraints of the photovoltaic source. An electrode rinse pump 609 may be provided separately to supply a dedicated flow stream to the electrode compartments, helping to remove gaseous products that form during the electrodialysis process. The electrode rinse pump 609 can prevent accumulation of hydrogen and oxygen gases at the electrode surfaces 316, 318 that could otherwise reduce system efficiency or create operational instabilities.

[0117] Valves 604 for flow control may be positioned throughout the hydraulic circuit to enable adjustment of flow rates and flow distribution between different channels within the electrodialysis stack 300. The valve arrangement can provide manual or automated control of the hydraulic conditions that affect ion transport rates and desalination performance. The flow control valves may enable optimization of the flow distribution to maintain equivalent flow rates through the diluate and concentrate channels, supporting balanced operation and consistent product water quality.

[0118] The electrodialysis stack 300 may be centrally positioned within the system as the primary water processing component, with power to the electrodes supplied directly from the control board through dedicated electrical connections. The PC Cell 64002 electrodialysis stack can include 20 cell pairs arranged in a compact configuration suitable for home-scale water treatment applications. The 20 cell pair configuration may provide sufficient membrane surface area for effective ion separation while maintaining a compact form factor appropriate for residential installations. The stack design can accommodate the voltage and current levels provided by the multi-converter MPPT system while delivering the desalination performance for successful brackish water treatment.

[0119] The system 600 may include an ion exchange membrane unit that houses the selectively permeable membranes responsible for ionic separation during the electrodialysis process, as mentioned above. For example, the membrane unit can contain alternating cation exchange membranes 308 and anion exchange membranes 306 arranged to create the diluate and concentrate flow channels. An exit cup may be provided for processed water collection, enabling measurement of product water flow rates and quality parameters. The exit cup arrangement can facilitate sampling of the desalinated water for conductivity measurements and other quality assessments during experimental operation.Attorney Docket No.: MIT 25641 USPCT | 88212-431989

[0120] Referring to Figs. 6A-6F, the experimental results demonstrate the performance characteristics of the multi-converter MPPT system operating with an electrodialysis desalination setup under controlled step response conditions. The experimental validation may utilize a Chroma 62120D-1200 combined power supply and solar simulator to mimic the procedure detailed in EN 50530:2010 Sequence A, which examines overall efficiency by stepping between minimum and maximum power available levels. The solar simulator can enable direct control of voltage maximum power point, maximum power output, and irradiance parameters to generate equivalent photovoltaic array characteristics including open circuit voltage, maximum power voltage, short circuit current, and maximum power current.

[0121] The step response testing methodology can involve successive transitions between two simulated solar panel profiles in approximately two-minute intervals to create controlled variations in available power and optimal operating conditions. The testing approach may maintain constant maximum power output while varying the voltage and current characteristics to simulate environmental changes such as temperature variations or irradiance fluctuations that would occur in real-world photovoltaic installations. The controlled step response conditions can enable systematic evaluation of the system's dynamic response characteristics and the effectiveness of the integrated control algorithms.

[0122] FIG. 6 A illustrates the voltage tracking performance of the multi-converter MPPT systems 104, 204, 404 of the present embodiments , demonstrating the system's ability to identify and track the maximum power point voltage under varying conditions. The voltage response data may show7three distinct traces: V P (A) representing the maximum power point voltage setpoint that steps between approximately 20V and 23V, Vsetpoint (B) showing the controller's voltage setpoint as determined by the MPPT algorithm, and VPV (C) depicting the actual voltage drawn from the simulated photovoltaic array. The shaded regions of FIG. 6A can distinguish between voltage step increases and decreases to facilitate visual comparison of system response characteristics.

[0123] The voltage tracking results can demonstrate that the MPPT algorithm successfully identifies the voltage at the maximum power point, with the controller voltage setpoint Vsetpoint eventually tracking the maximum power point voltage VMP through algorithmic adjustment. The actual simulated solar panel voltage VPV may respond to track the desired control setpoint Vsetpoint, indicating that the downstream converters adjust appropriately to maintain the optimal operating point. The convergence of VPV to VMP can indicate that theAttorney Docket No.: MIT 25641 USPCT | 88212-431989 overall system successfully tracks the maximum available power point, validating the maximum power point tracking portion of the integrated controller.

[0124] FIG. 6B presents the current tracking behavior of the system, showing how the maximum available current varies inversely with voltage changes to maintain constant power levels during the step response testing. The current response data may include IMP (D) representing the maximum available current from the solar simulator that shifts up and down opposite to voltage changes, and Ipv (E) representing the actual current drawn by the system from the simulated solar panel. The maximum available current IMP can shift in a manner that maintains constant overall power level while the actual current drawn IPV responds by approaching and tracking the maximum available current.

[0125] The current tracking results can demonstrate that the system successfully draws the maximum available current at each operating point, with the convergence of Ipv to IMP indicating effective power extraction from the photovoltaic source. The current response characteristics may show that the system adapts to varying current availability while maintaining optimal power extraction performance. The combination of voltage tracking shown in FIG. 6A and current tracking shown in FIG. 6B can provide evidence that the system successfully implements a method of drawing pow er from a photovoltaic array while using a maximum power point tracker in communication with the PV array to regulate power delivered to multiple downstream loads.

[0126] FIG. 6C displays the flow rate stability throughout the electrodialysis system during the experimental testing period, demonstrating the hydraulic performance characteristics under varying power conditions. The flow rate measurements may be presented in mL / min and show' relatively stable flow' with minor perturbations during maximum pow er point transitions. The flow rate data can be smoothed using a moving average with a span of 1000 samples to reduce measurement noise and provide clear visualization of the underlying flow rate trends. The flow rate stabi 1 i ty can indicate that the hydraulic load maintains consistent operation despite variations in available photovoltaic power and changes in the maximum power point tracking setpoint.

[0127] The flow' rate results may demonstrate that the pump load operates effectively within the multi-converter MPPT architecture, maintaining adequate water circulation through the electrodialysis stack while adapting to power availability changes. The relatively constant flow rate can indicate that the system successfully optimizes performance of the plurality ofAttorney Docket No.: MIT 25641 USPCT | 88212-431989 loads using the plurality of power converters, with the hydraulic load representing one component of the multi-load system. The flow rate measurements can provide input data for the limiting current calculation model that determines optimal operating conditions for the electrodialysis stack load.

[0128] FIG. 6D shows conductivity measurements for both the feed water stream and the diluate output stream, providing information about water quality characteristics and desalination performance throughout the experimental testing period. The conductivity data may be presented in pS / cm with the feed conductivity remaining relatively constant while the diluate conductivity shows variations related to the applied current and flow rate conditions. The feed water conductivity can remain stable because the diluate and concentrate streams are remixed to reconstitute the feed water in the continuous operation testing configuration.

[0129] The conductivity measurements can demonstrate that the electrodialysis stack load operates effectively as part of the multi-converter system, producing desalinated water with reduced ionic concentration compared to the feed water. The diluate conductivity fluctuations may be closely related to the flow rate and current applied to the electrodialysis stack, with greater applied current producing proportionally lower output conductivity. The conductivity data can provide evidence that the system implements a method where the plurality of loads comprises an electrodialysis stack that operates under optimal conditions while the maximum power point of the PV array is tracked simultaneously.

[0130] FIG. 6E presents the stack current control performance, illustrating three related traces that demonstrate the effectiveness of the feedback control system for maintaining optimal electrodialysis operation. The stack current control data may include the stack biasing coefficient a shown on the right axis (F), the applied current to the electrodialysis stack (G), and the commanded current setpoint (H) derived from the limiting current calculation based on flow rate and conductivity measurements. The commanded current can remain roughly constant throughout the experiment because the flow rate and diluate conductivity remain relatively stable, with exceptions for perturbations influenced by initial conductivity changes and occasional sensor noise.

[0131] The stack current control results can demonstrate that the system successfully uses a feedback loop to adjust a weighting factor on a duty cycle on a power converter of the plurality of power converters. The current command setpoint may be achieved by altering the stack biasing factor a, which in turn affects the stack power converter's duty cycle throughAttorney Docket No.: MIT 25641 USPCT | 88212-431989 the weighting mechanism. The stack biasing factor can fluctuate in response to MPPT voltage setpoint changes, while the applied current correspondingly matches the commanded current once settled, indicating that the downstream converter accomplishes its individual objective of maximizing and tracking limiting current while simultaneously adapting to maximum power point changes.

[0132] The stack current regulation can provide evidence that the system optimizes performance of the plurality of loads by optimizing power delivered to each of the plurality' of loads through independent control mechanisms. The electrodialysis stack current control may represent the applied electrical load component of the system, while the pump operation represents the hydraulic load component. The non-linear relationship between these loads can be accommodated through the weighting factor adjustment mechanism that enables independent optimization of each load while maintaining coordination through the centralized MPPT algorithm.

[0133] FIG. 6F illustrates the output load voltages and duty cycles for both converters, demonstrating the independent control capabilities of the multi-converter architecture. The converter control data may include solid lines representing the pump voltage Vpump and stack voltage Vstack, and dashed lines showing the corresponding duty cycles Dpump and Dstack. The duty cycle signals can directly affect the voltages delivered to the pump and stack loads, with both duty cycles changing in response to alterations in the maximum power point voltage setpoint.

[0134] The independent duty cycle control results can demonstrate that the two downstream converters maintain independent load control while still tracking the maximum power point of the photovoltaic array. The duty7cycles may show that perturbations in the maximum power point consistently cause initial drops in the commanded duty for both the pump and stack, followed by appropriate increases as optimal voltages are established. The independent duty7cycle operation can provide evidence that the power delivered to each of the plurality7of loads and the maximum power point of the PV array are tracked simultaneously through the coordinated control architecture.

[0135] The experimental results can collectively demonstrate that the system successfully implements a method of driving a plurality of loads where a first load of the plurality of loads is a hydraulic load and a second load of the plurality7of loads is an applied electrical load.The pump operation may represent the hydraulic load that provides water circulation throughAttorney Docket No.: MIT 25641 USPCT | 88212-431989 the electrodialysis system, while the electrodialysis stack operation represents the applied electrical load that performs the ion separation and desalination function. The plurality’ of loads can be non-linearly related through their interdependent effects on system performance, with pump flow rate affecting the limiting current calculation for the electrodialysis stack and stack current affecting the overall power consumption that influences pump operation.

[0136] The validation testing can demonstrate that the power drawn from the PV array is not buffered by a battery and that the power from the PV array does not pass through a converter upstream of the MPPT. The direct-drive operation may be evidenced by the immediate response of both loads to changes in photovoltaic array conditions without the buffering effects that would be present with intermediate energy storage. The elimination of upstream power conversion stages can be demonstrated by the direct connection between the photovoltaic simulator and the MPPT system, with power flowing directly to the downstream converters without intermediate conditioning or storage.

[0137] The simultaneous achievement of maximum power point tracking and optimal desalination operation can be validated through the coordinated response of all system parameters to the step changes in photovoltaic array conditions. The voltage and current tracking performance may demonstrate effective maximum power extraction, while the stable flow rate, appropriate conductivity reduction, and regulated stack current demonstrate optimal desalination operation. The independent duty cycle control can enable each load to maintain its optimal operating conditions while contributing to overall system performance, validating the effectiveness of the multi-converter MPPT architecture for direct-drive renewable energy applications.

[0138] The multi-converter maximum power point tracker system architecture can be adapted to accommodate a wide range of applications beyond electrodialysis desalination, providing direct-drive renewable energy solutions for various load types that can operate at variable power levels. The fundamental control principles and hardware architecture may be generalized to support diverse applications where the elimination of energy storage provides advantages in terms of cost reduction, system simplification, and environmental impact minimization.

[0139] Water processing systems can benefit from the multi-converter MPPT architecture through the coordination of multiple pumps, treatment stages, and monitoring equipment within integrated water treatment facilities. The system architecture may accommodateAttorney Docket No.: MIT 25641 USPCT | 88212-431989 reverse osmosis systems where high-pressure pumps and membrane cleaning systems use coordinated power management to optimize water production while maintaining membrane performance. Ultrafiltration and microfiltration systems can utilize the multiconverter approach to balance power between circulation pumps, backwash systems, and chemical dosing pumps based on membrane fouling conditions and treatment parameters. The plurality of loads can comprise water processing systems that include multiple pumping stages, chemical injection systems, and monitoring equipment that operate with variable power parameters based on water quality conditions and treatment objectives.

[0140] Filtration systems can implement the multi-converter MPPT architecture to coordinate power distribution between filtration pumps, backwash systems, and regeneration equipment in applications such as sand filtration, activated carbon treatment, and ion exchange systems. The control system may optimize power allocation between forward filtration operation and periodic backwash cycles based on pressure differential measurements and filtration efficiency indicators. Multi-stage filtration systems can benefit from independent power regulation for each filtration stage while maintaining overall system coordination through the centralized MPPT algorithm. The variable power operation capability may enable filtration systems to adapt their treatment intensity based on available solar power while maintaining minimum treatment standards through load prioritization strategies.

[0141] Waste processing systems can utilize the multi-converter architecture to coordinate power distribution among various treatment processes including aeration systems, mixing equipment, and separation devices in wastewater treatment applications. The system architecture may support activated sludge processes where aeration blowers and return sludge pumps use coordinated operation to maintain proper biological treatment conditions. Anaerobic digestion systems can benefit from coordinated control of mixing systems, heating elements, and gas handling equipment to optimize biogas production while maintaining stable digester operation. The plurality of loads can comprise waste processing systems that include biological treatment processes, physical separation equipment, and chemical treatment systems that operate with interdependent power parameters.

[0142] Management of dual or multi-pumping architectures can be facilitated through the multi-converter MPPT system where multiple pumps operate in parallel or series configurations to achieve specific flow' and pressure parameters. The system architecture may coordinate pump operation in water distribution systems where multiple pumps provideAttorney Docket No.: MIT 25641 USPCT | 88212-431989 redundancy and variable capacity based on demand conditions. Irrigation systems can benefit from coordinated pump control where different zones use independent flow control while maintaining overall system pressure and flow balance. The plurality of loads can comprise dual or multi-pumping architectures where pump staging, speed control, and flow distribution use coordinated power management to optimize hydraulic performance and energy efficiency.

[0143] Solar HVAC systems can implement the multi-converter MPPT architecture to coordinate power distribution among various heating, ventilation, and air conditioning components that operate with variable power parameters based on thermal loads and environmental conditions. Chiller systems may utilize the multi-converter approach to balance power between compressors, circulation pumps, and cooling tower fans based on cooling demand and ambient conditions. The system architecture can optimize chiller operation by coordinating compressor staging with auxiliary equipment operation to maximize cooling efficiency while adapting to available solar power. Hot water heater systems can benefit from coordinated control of heating elements, circulation pumps, and temperature control systems to optimize thermal energy storage and distribution based on hot water demand patterns and solar availability.

[0144] The plurality of loads can comprise solar HVAC systems such as chillers and hot water heaters that coordinate power management among multiple components with different thermal time constants and operational priorities. Heat pump systems may utilize the multiconverter architecture to coordinate power between compressor operation, auxiliary heating elements, and circulation systems based on thermal load parameters and coefficient of performance optimization. Solar cooling systems can implement coordinated control of absorption chillers, circulation pumps, and heat rejection systems to optimize cooling production while maintaining stable operation under varying solar conditions.

[0145] Refrigeration and freezing units can benefit from the multi-converter MPPT architecture through coordinated control of compressors, evaporator fans, condenser fans, and defrost systems that operate with different thermal parameters and operational schedules. The system architecture may enable independent temperature control for multiple refrigeration zones while optimizing overall energy consumption based on available solar power. Commercial refrigeration systems can utilize the multi-converter approach to coordinate operation of multiple compressor stages, case lighting, and anti-sweat heaters based on thermal loads and energy availability. The plurality of loads can includeAttorney Docket No.: MIT 25641 USPCT | 88212-431989 refrigeration and freezing units that include multiple temperature zones, defrost systems, and auxiliary equipment that use independent control while maintaining coordination for optimal energy utilization. ro 1461 Walk-in coolers and freezers may implement the multi-converter architecture to balance power between refrigeration compressors, evaporator fans, and door heaters based on usage patterns and thermal load variations. Ice making systems can benefit from coordinated control of refrigeration equipment, water pumps, and harvest systems to optimize ice production while adapting to available solar power. The variable power operation capability may enable refrigeration systems to implement thermal storage strategies where excess solar power is used for additional cooling during peak availability periods.

[0147] Charging systems for fleets of electric vehicles can utilize the multi-converter MPPT architecture to coordinate power distribution among multiple charging stations while optimizing power allocation based on vehicle charging parameters and grid interaction capabilities. The system architecture may enable dynamic load balancing among multiple electric vehicle charging points to maximize charging throughput while maintaining power quality and system stability. Fleet charging systems can benefit from coordinated control where charging power is allocated based on vehicle battery state of charge, departure schedules, and available renewable energy. The plurality of loads can comprise charging systems for fleets of electric vehicles that include multiple charging stations, power conditioning equipment, and energy management systems that implement coordinated operation to optimize charging efficiency and grid integration.

[0148] DC fast charging systems may implement the multi-converter approach to coordinate power between multiple charging ports while managing thermal loads from power electronics cooling systems. The system architecture can optimize charging station operation bybalancing power among active charging sessions while maintaining thermal management and power quality parameters. Solar-powered charging stations can benefit from direct-drive operation that eliminates battery storage costs while providing coordinated charging services based on solar availability and vehicle charging priorities.

[0149] The control system architecture can be adapted to accommodate different load characteristics through modification of the weighting factors, feedback control parameters, and optimization objectives while maintaining the fundamental multi-converter MPPT framework. Each application may use specific sensor inputs, control algorithms, andAttorney Docket No.: MIT 25641 USPCT | 88212-431989 operational constraints that can be implemented within the generalized control structure. The system architecture provides a control system including a maximum power point tracker configured to operate with various load types that can benefit from direct-drive renewable energy operation without intermediate energy storage.

[0150] The adaptability of the multi-converter MPPT system may enable deployment across diverse applications where multiple loads implement coordinated power management while maintaining optimal renewable energy utilization. The elimination of energy storage parameters can provide cost and complexity advantages across all these applications while enabling responsive load management that adapts to varying renewable energy availability. The control system architecture can accommodate load-specific optimization objectives while maintaining the core functionality of maximum power point tracking and coordinated power distribution among multiple variable loads.

[0151] Referring to FIG. 8, a computing system architecture illustrates the hardware and software components that can be used to implement the control algorithms for the multiconverter maximum power point tracker system. The computing system may provide the computational resources useful for executing the integrated MPPT and optimal desalination control algorithms while supporting advanced control implementations through machine learning and distributed computing capabilities. The computing architecture can enable sophisticated control strategies that extend beyond basic feedback control to incorporate predictive algorithms, adaptive control parameters, and distributed monitoring capabilities.

[0152] FIG. 8 is a schematic diagram that shows an example of a computing system 1600 that can be used to implement the techniques described herein. The computing system 1600 includes one or more computing devices (e.g, computing device 1610), which can be in wired and / or wireless communication with various peripheral device(s) 1680, data source(s) 1690, and / or other computing devices (e.g., over network(s) 1670). The computing device 1610 can represent various forms of stationary computers 1616 (e.g., workstations, kiosks, servers, mainframes, edge computing devices, quantum computers, etc.) and mobile computers 1614 (e.g., laptops, tablets, mobile phones, personal digital assistants, wearable devices, etc.). In some implementations, the computing device 1610 can be included in (and / or in communication with) various other sorts of devices, such as data collection devices (e.g., devices that are configured to collect data from a physical environment, such as microphones, cameras, scanners, sensors, etc.), robotic devices (e.g., devices that are configured to physically interact with objects in a physical environment, such asAttorney Docket No.: MIT 25641 USPCT | 88212-431989 manufacturing devices, maintenance devices, object handling devices, etc.), vehicles (e.g., devices that are configured to move throughout a physical environment, such as automated guided vehicles, manually operated vehicles, etc.), or other such devices. Each of the devices (e.g., stationary computers, mobile computers, and / or other devices) can include components of the computing device 1610, and an entire system can be made up of multiple devices communicating with each other. For example, the computing device 1610 can be part of a computing system that includes a network of computing devices, such as a cloud-based computing system, a computing system in an internal network, or a computing system in another sort of shared network. Processors of the computing device 1610 and other computing devices of a computing system can be optimized for different types of operations, secure computing tasks, etc. The components shown herein, and their functions, are meant to be examples, and are not meant to limit implementations of the technology described and / or claimed in this document.

[0153] The computing device 1610 includes processor(s) 1620, memory device(s) 1630, storage device(s) 1640, and interface(s) 1650. Each of the processor(s) 1620, the memory device(s) 1630, the storage device(s) 1640, and the interface(s) 1650 are interconnected using a system bus 1660. The processor(s) 1620 are capable of processing instructions for execution within the computing device 1610, and can include one or more single-threaded and / or multi-threaded processors. The processor(s) 1620 are capable of processing instructions stored in the memory7device(s) 1630 and / or on the storage device(s) 1640. The memory device(s) 1630 can store data within the computing device 1610, and can include one or more computer-readable media, volatile memory units, and / or non-volatile memory units. The storage device(s) 1640 can provide mass storage for the computing device 1610, can include various computer-readable media (e.g., a floppy disk device, a hard disk device, a tape device, an optical disk device, a flash memory or other similar solid state memory device, or an array of devices, including devices in a storage area network or other configurations), and can provide date security / encryption capabilities.

[0154] The interface(s) 1650 can include various communications interfaces (e.g., USB, Near-Field Communication (NFC), Bluetooth, WiFi, Ethernet, wireless Ethernet, etc.) that can be coupled to the network(s) 1670, peripheral device(s) 1680, and / or data source(s) 1690 (e.g., through a communications port, a network adapter, etc.). Communication can be provided under various modes or protocols for wired and / or wireless communication. Such communication can occur, for example, through a transceiver using a radio-frequency. AsAttorney Docket No.: MIT 25641 USPCT | 88212-431989 another example, communication can occur using light (e.g, laser, infrared, etc.) to transmit data. As another example, short-range communication can occur, such as using Bluetooth, WiFi, or other such transceiver. In addition, a GPS (Global Positioning System) receiver module can provide location-related wireless data, which can be used as appropriate by device applications. The interface(s) 1650 can include a control interface that receives commands from an input device (e.g, operated by a user) and converts the commands for submission to the processors 1620. The interface(s) 1650 can include a display interface that includes circuitry for driving a display to present visual information to a user. The interface(s) 1650 can include an audio codec which can receive sound signals (e.g, spoken information from a user) and convert it to usable digital data. The audio codec can likewise generate audible sound, such as through an audio speaker. Such sound can include real-time voice communications, recorded sound (e.g, voice messages, music files, etc.), and / or sound generated by device applications.

[0155] The network(s) 1670 can include one or more wired and / or wireless communications netw orks, including various public and / or private networks. Examples of communication networks include a LAN (local area network), a WAN (wide area network), and / or the Internet. The communication networks can include a group of nodes (e.g, computing devices) that are configured to exchange data (e.g, analog messages, digital messages, etc.), through telecommunications links. The telecommunications links can use various techniques (e.g, circuit switching, message switching, packet switching, etc.) to send the data and other signals from an originating node to a destination node. In some implementations, the computing device 1610 can communicate with the peripheral device(s) 1680. the data source(s) 1690, and / or other computing devices over the network(s) 1670. In some implementations, the computing device 1610 can directly communicate with the peripheral device(s) 1680, the data source(s), and / or other computing devices.

[0156] The peripheral device(s) 1680 can provide input / output operations for the computing device 1610. Input devices (e.g., keyboards, pointing devices, touchscreens, microphones, cameras, scanners, sensors, etc.) can provide input to the computing device 1610 (e.g, user input and / or other input from a physical environment). Output devices (e.g., display units such as display screens or projection devices for displaying graphical user interfaces (GUIs)), audio speakers for generating sound, tactile feedback devices, printers, motors, hardware control devices, etc.) can provide output from the computing device 1610 (e.g.. user-directed output and / or other output that results in actions being performed in a physicalAttorney Docket No.: MIT 25641 USPCT | 88212-431989 environment). Other kinds of devices can be used to provide for interactions between users and devices. For example, input from a user can be received in any form, including visual, auditory, or tactile input, and feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditor feedback, or tactile feedback).

[0157] The data source(s) 1690 can provide data for use by the computing device 1610, and / or can maintain data that has been generated by the computing device 1610 and / or other devices (e.g., data collected from sensor devices, data aggregated from various different data repositories, etc.). In some implementations, one or more data sources can be hosted by the computing device 1610 (e.g., using the storage device(s) 1640). In some implementations, one or more data sources can be hosted by a different computing device. Data can be provided by the data source(s) 1690 in response to a request for data from the computing device 1610 and / or can be provided without such a request. For example, a pull technology can be used in which the provision of data is driven by device requests, and / or a push technology can be used in which the provision of data occurs as the data becomes available (e.g., real-time data streaming and / or notifications). Various sorts of data sources can be used to implement the techniques described herein, alone or in combination.

[0158] In some implementations, a data source can include one or more data store(s) 1690a. The database(s) can be provided by a single computing device or network (e.g., on a file system of a server device) or provided by multiple distributed computing devices or networks (e.g., hosted by a computer cluster, hosted in cloud storage, etc.). In some implementations, a database management system (DBMS) can be included to provide access to data contained in the database(s) (e.g., through the use of a query language and / or application programming interfaces (APIs)). The database(s), for example, can include relational databases, object databases, structured document databases, unstructured document databases, graph databases, and other appropriate ty pes of databases.

[0159] In some implementations, a data source can include one or more blockchains 1690b. A blockchain can be a distributed ledger that includes blocks of records that are securely linked by cryptographic hashes. Each block of records includes a cryptographic hash of the previous block, and transaction data for transactions that occurred during a time period. The blockchain can be hosted by a peer-to-peer computer network that includes a group of nodes (e.g., computing devices) that collectively implement a consensus algorithm protocol to validate new transaction blocks and to add the validated transaction blocks to the blockchain. By storing data across the peer-to-peer computer network, for example, theAttorney Docket No.: MIT 25641 USPCT | 88212-431989 blockchain can maintain data quality (e.g., through data replication) and can improve data trust (e.g., by reducing or eliminating central data control).

[0160] In some implementations, a data source can include one or more machine learning systems 1690c. The machine learning system(s) 1690c. for example, can be used to analyze data from various sources (e.g, data provided by the computing device 1610, data from the data store(s) 1690a, data from the blockchain(s) 1690b, and / or data from other data sources), to identify patterns in the data, and to draw inferences from the data patterns. In general, training data 1692 can be provided to one or more machine learning algorithms 1694, and the machine learning algorithm(s) can generate a machine learning model 1696. Execution of the machine learning algorithm(s) can be performed by the computing device 1610, or another appropriate device. Various machine learning approaches can be used to generate machine learning models, such as supervised learning (e.g, in which a model is generated from training data that includes both the inputs and the desired outputs), unsupervised learning (e.g., in which a model is generated from training data that includes only the inputs), reinforcement learning (e.g, in which the machine learning algorithm(s) interact with a dynamic environment and are provided with feedback during a training process), or another appropriate approach. A variety of different types of machine learning techniques can be employed, including but not limited to convolutional neural networks (CNNs), deep neural networks (DNNs), recurrent neural networks (RNNs), and other types of multi-layer neural networks.

[0161] Various implementations of the systems and techniques described herein can be realized in digital electronic circuitry', integrated circuitry, specially designed ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. A computer program product can be tangibly embodied in an information carrier (e.g., in a machine-readable storage device), for execution by a programmable processor. Various computer operations (e.g., methods described in this document) can be performed by a programmable processor executing a program of instructions to perform functions of the described implementations by operating on input data and generating output. The described features can be implemented in one or more computer programs that are executable on a programmable system including at least one programmable processor coupled to receive data and instructions from, and to transmit data and instructions to, a data storage system, at least one input device, and at least one output device. A computer program is a set of instructions that can be used, directly or indirectly, by aAttorney Docket No.: MIT 25641 USPCT | 88212-431989 computer to perform a certain activity or bring about a certain result. A computer program can be written in any form of programming language, including compiled or interpreted languages, and can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program product can be a computer- or machine-readable medium, such as a storage device or memory device. As used herein, the terms machine-readable medium and computer-readable medium refer to any computer program product, apparatus and / or device (e.g., magnetic discs, optical disks, memory, etc.) used to provide machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term machine-readable signal refers to any signal used to provide machine instructions and / or data to a programmable processor.

[0162] Suitable processors for the execution of a program of instructions include, by way of example, both general and special purpose microprocessors, and can be a single processor or one of multiple processors of any kind of computer. Generally, a processor will receive instructions and data from a read-only memory or a random access memory or both. The elements of a computer are a processor for executing instructions and one or more memory devices for storing instructions and data. Generally, a computer can also include, or can be operatively coupled to communicate with, one or more mass storage devices for storing data files. Such devices can include magnetic disks (e.g., internal hard disks and / or removable disks), magneto-optical disks, and optical disks. Storage devices suitable for tangibly embodying computer program instructions and data can include all forms of non-volatile memory, including by way of example semiconductor memory devices, flash memory devices, magnetic disks (e.g., internal hard disks and removable disks), magneto-optical disks, and optical disks. The processor and the memory’ can be supplemented by, or incorporated in, ASICs (application-specific integrated circuits).

[0163] The systems and techniques described herein can be implemented in a computing system that includes a back end component (e.g., a data server), or that includes a middleware component (e.g. , an application server), or that includes a front end component (e.g. , a client computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end. middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g, a communication network). The computer system can include clients and servers, which canAttorney Docket No.: MIT 25641 USPCT | 88212-431989 be generally remote from each other and ty pically interact through a network, such as the described one. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. ro 1641 Examples of the above-described embodiments can include the following:1. A direct-drive system architecture, comprising: a photovoltaic (PV) array; a maximum power point tracker (MPPT) in communication with the PV array; a plurality of power converters located downstream of the MPPT and in communication with the MPPT and the PV array, the plurality of power converters being configured to receive a duty cycle command from the MPPT and at least one of a voltage and a current from the PV array; and a plurality of loads, with each load being dow nstream of, and in communication with, one or more of the plurality of pow er converters, with each of the plurality of power converters being configured to independently regulate one or more of the plurality of loads, w herein the MPPT is configured to regulate: (i) each of the dow nstream plurality of loads independently; and (ii) a power point of the PV array.2. The system architecture of example 1, wherein the plurality of power converters are variable direct current (DC) / DC converters.3. The system architecture of example 1 or example 2, wherein the plurality of loads and the power point of the PV array are regulated simultaneously.4. The system architecture of any of examples 1 to 3, further comprising a feedback loop that includes a weighting factor on a duty cycle on at least one of the plurality of pow er converters.5. The direct-drive system architecture of example 4, wherein the weighting factor is dynamically adjusted based on sensed process variables from the plurality of loads.6. The system architecture of any of examples 1 to 5, w herein the plurality of pow er converters comprise a dual buck converter.Attorney Docket No.: MIT 25641 USPCT | 88212-4319897. The direct-drive system architecture of any of examples 1 to 6, wherein the plurality of power converters operate in a continuous conduction mode with a cunent ripple constraint or a discontinuous conduction mode.8. The system architecture of any of examples 1 to 7, wherein the MPPT comprises a plurality of power supplies, with each power supply being in communication with a power converter of the plurality of power converters.9. The system architecture of any of examples 1 to 8, wherein each load associated with each of the plurality of power converters is configured to be operated simultaneously.10. The system architecture of any of examples 1 to 9, wherein the plurality of loads includes a hydraulic load and an applied electrical load.11. The system architecture of example 10, wherein the plurality of loads includes an electrodialysis (ED) stack or a pump.12. The system architecture of any of examples 1 to 9, wherein no converter exists upstream of the MPPT.13. The system architecture of any of examples 1 to 10, wherein the PV array and the MPPT are not in communication with a battery.14. The system architecture of any of examples 1 to 13, wherein the plurality of loads comprise one or more of water processing systems, filtration systems, waste processing systems, management of dual or multi-pumping architectures, solar HVAC systems (e.g., chillers, hot water heaters), refrigeration and freezing units, or charging system for fleets of electric vehicles.15. The system architecture of any of examples 1 to 14, wherein the MPPT is devoid of intermediate energy storage between the PV array and the plurality7of loads.16. A multi-converter maximum power point tracking system, comprising: a photovoltaic (PV) array configured to generate electrical power; a controller in communication with the PV array and configured to execute a maximum power point tracking (MPPT) algorithm;Attorney Docket No.: MIT 25641 USPCT | 88212-431989 a plurality' of variable direct current / direct current (DC / DC) converters downstream of the controller, each variable DC / DC converter configured to receive a weighted duty cycle signal from the controller; a plurality of loads, each load powered by a respective variable DC / DC converter; and a feedback control system configured to adjust weighting factors for the duty cycle signals based on individual load requirements, wherein the controller is configured to coordinate power distribution among the plurality of loads while maintaining maximum power extraction from the PV array.17. The multi-converter maximum power point tracking system of example 16, wherein the maximum power point tracking algorithm comprises a perturb and observe algorithm.18. The multi-converter maximum power point tracking system of example 16 or example 17, wherein the plurality7of loads comprises an electrodialysis stack and a pump system.19. The multi-converter maximum power point tracking system of example 18, wherein the electrodialysis stack comprises alternating cation exchange membranes and anion exchange membranes arranged between diluate and concentrate flow channels.20. The multi-converter maximum power point tracking system of example 19, wherein the feedback control system is configured to maintain operation of the electrodialysis stack at or below a limiting current threshold based on measured flow7rate and conductivity' parameters.21. A method of driving a plurality’ of loads, comprising: drawing power from a photovoltaic (PV) array; and using a maximum pow er point tracker (MPPT) in communication with the PV array and a plurality7of power converters disposed downstream of the MPPT to regulate a power delivered to one or more loads located downstream of the plurality of power converters.22. The method of example 21, further comprising optimizing performance of the one or more loads using the plurality7of pow er converters.Attorney Docket No.: MIT 25641 USPCT | 88212-43198923. The method of example 22, wherein optimizing performance further comprises optimizing a power delivered to each of the plurality of loads.24. The method of any of examples 21 to 23, further comprising tracking the power delivered to each of the plurality of loads and a maximum power point of the PV array simultaneously.25. The method of any of examples 21 to 24, wherein the plurality7of loads comprises an electrodialysis (ED) stack.26. The method of any of examples 21 to 25, wherein the plurality7of loads are non- linearly related.27. The method of example 26, wherein a first load of the plurality of loads is a hydraulic load and a second load of the plurality of loads is an applied electrical load.28. The method of any7of examples 21 to 27, further comprising using a feedback loop to adjust a weighting factor on a duty7cycle on a power converter of the plurality7of power converters.29. The method of any of examples 21 to 28, wherein the power drawn from the PV array is not buffered by a battery.30. The method of any of examples 21 to 29, wherein the power from the PV array does not pass through a converter upstream of the MPPT.31. A method of operating multiple loads from a photovoltaic source, comprising: drawing power from a photovoltaic (PV) array using a maximum power point tracker(MPPT); generating a common duty cycle command based on maximum power point tracking of the PV array; applying individual weighting factors to the common duty cycle command to create weighted duty7cycle signals for a plurality' of power converters; delivering the weighted duty cycle signals to the plurality of power converters to independently regulate power to a plurality of loads; andAttorney Docket No.: MIT 25641 USPCT | 88212-431989 adjusting the weighting factors based on feedback from the plurality of loads to optimize performance of each load while simultaneously tracking the maximum power point of the PV array.32. The method of example 31, wherein the plurality of loads comprises an electrodialysis stack and a pump system.33. The method of example 31 or example 32, wherein adjusting the weighting factors comprises maintaining operation of a load of the plurality of loads at or below a limiting current threshold based on measured flow rate and conductivity parameters.34. The method of example 33, wherein the limiting current threshold is calculated using an empirical model that relates limiting current to solution conductivity and flow rate.

[0165] One skilled in the art will appreciate further features and advantages of the disclosure based on the above-described embodiments. Accordingly, the disclosure is not to be limited by what has been particularly shown and described, except as indicated by the appended claims. By way of example, the multi -converter MPPT systems described herein can be adapted for use with various renewable energy applications including solar-powered water treatment systems, off-grid industrial processes, electric vehicle charging stations, and distributed energy management systems where multiple loads implement independent control while maximizing power extraction from photovoltaic sources. A person skilled in the art, in view of the present disclosures, will be able to adapt some or all of the various systems, devices, and methods disclosed herein for other electrochemical processes such as electrolysis for hydrogen production, electrowinning operations, carbon dioxide reduction systems, and various pumping applications in agriculture, aquaculture, and remote water supply systems where direct-drive operation without energy' storage is advantageous. All publications and references cited herein are expressly incorporated herein by reference in their entirety.

Claims

Attorney Docket No.: MIT 25641 USPCT | 88212-431989CLAIMSWhat is claimed is:

1. A direct-drive system architecture, comprising: a photovoltaic (PV) array; a maximum power point tracker (MPPT) in communication with the PV array; a plurality of power converters located downstream of the MPPT and in communication with the MPPT and the PV array, the plurality of power converters being configured to receive a duty cycle command from the MPPT and at least one of a voltage and a current from the PV array; and a plurality of loads, with each load being downstream of, and in communication with, one or more of the plurality' of power converters, with each of the plurality of power converters being configured to independently regulate one or more of the plurality of loads, wherein the MPPT is configured to regulate: (i) each of the dow nstream plurality of loads independently; and (ii) a power point of the PV array.

2. The system architecture of claim 1, wherein the plurality of power converters are variable direct current (DC) / DC converters.

3. The system architecture of claim 1, wherein the plurality of loads and the pow er point of the PV array are regulated simultaneously.

4. The system architecture of claim 1, further comprising a feedback loop that includes a w eighting factor on a duty cycle on at least one of the plurality of power converters.

5. The direct-drive system architecture of claim 4, wherein the weighting factor is dynamically adjusted based on sensed process variables from the plurality' of loads.

6. The system architecture of claim 1, wherein the plurality of power converters comprise a dual buck converter.

7. The direct-drive system architecture of claim 1, wherein the plurality of power converters operate in a continuous conduction mode with a current ripple constraint or a discontinuous conduction mode.Attorney Docket No.: MIT 25641 USPCT | 88212-4319898. The system architecture of claim 1, wherein the MPPT comprises a plurality of power supplies, with each power supply being in communication with a power converter of the plurality of power converters.

9. The system architecture of claim 1, wherein the plurality of loads includes a hydraulic load and an applied electrical load.

10. The system architecture of claim 9, wherein the plurality of loads includes an electrodialysis (ED) stack or a pump.1 1. The system architecture of claim 1, wherein the MPPT is devoid of intermediate energy storage between the PV array and the plurality of loads.

12. A method of driving a plurality of loads, comprising: drawing power from a photovoltaic (PV) array: and using a maximum power point tracker (MPPT) in communication with the PV array and a plurality of power converters disposed downstream of the MPPT to regulate a power delivered to one or more loads located downstream of the plurality of power converters.

13. The method of claim 12, further comprising tracking the power delivered to each of the plurality of loads and a maximum power point of the PV array simultaneously.

14. The method of claim 12. wherein the plurality of loads are non-linearly related.

15. The method of claim 14, wherein a first load of the plurality of loads is a hydraulic load and a second load of the plurality of loads is an applied electrical load.

16. The method of claim 12, further comprising using a feedback loop to adjust a weighting factor on a duty cycle on a power converter of the plurality of power converters.

17. The method of claim 12, wherein the power drawn from the PV array is not buffered by a batten’.

18. A method of operating multiple loads from a photovoltaic source, comprising:Attorney Docket No.: MIT 25641 USPCT | 88212-431989 drawing power from a photovoltaic (PV) array using a maximum power point tracker (MPPT); generating a common duty cycle command based on maximum power point tracking of the PV array; applying individual weighting factors to the common duty cycle command to create weighted duty cycle signals for a plurality of power converters; delivering the weighted duty cycle signals to the plurality of power converters to independently regulate power to a plurality of loads; and adjusting the weighting factors based on feedback from the plurality' of loads to optimize performance of each load while simultaneously tracking the maximum power point of the PV array.

19. The method of claim 18, wherein adjusting the weighting factors comprises maintaining operation of a load of the plurality' of loads at or below a limiting current threshold based on measured flow rate and conductivity parameters.

20. The method of claim 19, wherein the limiting current threshold is calculated using an empirical model that relates limiting current to solution conductivity and flow rate.

Images