Method for charging a battery based on nonlinear voltammetry and fast charging system implementing the method

Abstract

The present invention relates to a method based on nonlinear voltammetry (NLV) for charging a battery. The present invention also relates to a fast charging system implementing the method. Adaptive charging, nonlinear voltage change and relaxation are the key foundations of the method. Adaptive charging allows the system to balance the charge based on the user's time requirements, the required charge capacity and the SOC and SOH of the battery. Nonlinearly changing the voltage coupled with a suitable relaxation pattern allows the method to obtain the maximum charge capacity without damaging the battery.

Description

[0001] This patent application claims priority to Singapore patent application number 10201710151Y filed on December 7, 2017. Technical Field

[0002] The present invention relates to a method for charging a battery based on nonlinear voltammetry (NLV). The present invention also relates to a fast charging system for implementing the method. Background Art

[0003] “How to charge batteries faster?” is a question that has not been fully solved for decades since the advent of battery storage devices. More importantly, faster charging of lithium-ion batteries has become a key issue due to the urgent and massive use of mobile device technology and the growing demand for electric vehicles (EV) and plug-in hybrid vehicles (PHEV) in recent years (due to the urgent need to curb air pollution caused by petroleum-based vehicles). Therefore, fast charging solutions for lithium-ion batteries are a multi-billion dollar innovation in the world today.

[0004] The object of the present invention is to propose a new non-linear voltammetry (NLV) based charging protocol that allows fast charging of batteries with improved performance compared to the current constant current constant voltage (CCCV) fast charging technology. Summary of the invention

[0005] According to one aspect of the present invention, a method for charging a battery system includes:

[0006] a. Measure the voltage of multiple battery systems, 'v'

[0007] b. Measure the charging current of multiple battery systems, 'i'

[0008] c. Measure the temperature of multiple battery systems, 'T'

[0009] d. Measure the charging time, t ch

[0010] e. Measure the state of charge, SOC, of ​​multiple battery systems

[0011] f. Measure the number of charge and discharge cycles, 'n'

[0012] g. Apply the following relationship:

[0013]

[0014] in,

[0015] - is the time growth rate of voltage (Vs -1)

[0016] - is the absolute value of the charging current rate (mA.s -1 )

[0017] -K is a variable parameter, 0.006≤K≤300

[0018] -α is an adjustable constant, 0.01≤α≤100.

[0019] According to another aspect of the present invention, a battery charging system is provided, the battery charging system comprising:

[0020] h. a sensor for measuring the voltage 'v' of a plurality of battery systems,

[0021] i. a sensor for measuring the current 'i' of a plurality of battery systems,

[0022] j. a sensor for measuring the temperature 'T' of a plurality of battery systems,

[0023] k. A sensor for measuring the voltage 'v' of a plurality of battery systems,

[0024] l. Used to measure the charging time of multiple battery systems't ch 'Sensor,

[0025] m. A computing system for measuring the state of charge SOC of a battery system,

[0026] n. A computing system for measuring the number of charge and discharge cycles of a battery system,

[0027] o. A computing system for applying a relationship such as:

[0028]

[0029] in,

[0030] - is the time growth rate of voltage (Vs -1 )

[0031] - is the absolute value of the charging current rate (mA.s -1 )

[0032] -K is a variable parameter, 0.006≤K≤300

[0033] -α is an adjustable constant, 0.01≤α≤100.

[0034] The battery system may include one battery cell or a multi-battery cell system, and may be arranged in series and / or parallel battery cell configurations.

[0035] The voltage of the battery cell is, for example, between 2V and 5V, and the charging current in the battery cell is between 0C and 10C (the nC rate is defined so that the full charge time can be within 1000mA). The constant charging current in hours, that is, at a 10C rate, the charging time is hours = 6 minutes).

[0036] The temperature T of the battery cell can be between -20°C and +55°C, and the charging time t from 0% SOC to 100% SOC ch Between 10 minutes and 2 hours. The SOC can be between 0% and 100%, and the number of cycles is 200≤n≤2000.

[0037] An adaptive charging protocol (ACP) based on nonlinear voltammetry (NLV) for fast charging of lithium-ion batteries was developed to charge the battery in about 10 minutes. This is a combination of two fast charging methods that can be applied to any type of battery. The protocol is used as a memoryless charging model as well as a memory-based charging model. If historical data on the battery chemistry is available, this protocol automatically adjusts to use them to provide the best charging performance.

[0038] If a random battery happens to be charged, without any historical or specific data, a fast learned model about its ΔSOC will be fair enough to charge it quickly and safely. Not only that, the user's requirements and some system requirements (as and when these are detected) will also be taken into account when adjusting the protocol for charging. So this can also be considered as a general protocol for fast charging of batteries.

[0039] Using this method, the battery can be fully charged in about 10 minutes. In general, the battery will be charged in about 22 minutes to 24 minutes. Through cycle tests, it has been demonstrated that this charging protocol does not have a great impact on capacity fade. Moreover, this could be a model for fast charging of any type of battery, since the basis of this protocol is to let the battery charge with its own favorable current at any point in time depending on its ΔSOC and SOH.

[0040] Adaptive charging, nonlinear voltage change, and relaxation are the key foundations of the protocol. Adaptive charging allows the system to balance the charge based on the user's timing requirements, the required charge capacity, and the battery's SOC and SOH. Nonlinearly changing the voltage coupled with an appropriate relaxation pattern allows the method to achieve maximum charge capacity without straining the battery. The protocol uses high-speed NLV steps or configurable constant current (CC) charging at the starting SOC as the battery cell impedance increases toward end-of-discharge (EOD) [1]. If the system cannot achieve the expected charge at NLV-based charge termination, the adaptive protocol will decide whether to use another CC charge to achieve balanced capacity. The following summarizes NLV charging:

[0041] The NLV-based charging protocol of the present invention can also be used in combination with other fast charging protocols, such as the constant current protocol (CC), constant current constant voltage protocol (CCCV) and cascade pulse charging protocol described in PCT application #PCTIB2018 / 059705. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] These and other features and advantages of the present invention will become better understood with regard to the following description, appended claims and accompanying drawings, in which:

[0043] - Figure 1 is a functional schematic diagram of a fast charging system implementing an adaptive charging protocol according to the present invention,

[0044] - Figure 2 The voltage and current curves of the NLV-based charging process are illustrated.

[0045] - Figure 3 Illustrate the ACP-NLV charging protocol: process flow,

[0046] - Figure 4 is the possible range of initial SOC,

[0047] - Figure 5 This example shows a section of current change.

[0048] - Figure 6 This illustrates the process of training K and StepTime.

[0049] - Figure 7 exemplifies the AK value curve during the initial portion of charging, during the 1st (1 to 2) minute [at approximately 0% SOC],

[0050] - Figure 8 illustrates the AK value curve during the initial portion of charging, during (11 to 16) minutes [at approximately 40% SOC],

[0051] - Fig. 9 exemplifies the AK value curve during the initial portion of charging, during (23-25) minutes [at approximately 95% SOC],

[0052] - Fig.10 Indicates the process flow of controlling the "termination voltage",

[0053] - Fig.11 represents the voltage and current curves of random test 1,

[0054] - Fig.12 represents the voltage and current curves of random test 1,

[0055] - Fig.13 Illustrated is the NLV charging of Random Test 1: Current vs. K value during NLV charging ending with a full charge to 660 mAh in only 23.3 minutes,

[0056] - Fig.14 Illustrated is the NLV charging of Random Test 2: Current vs. K value during NLV charging ending with a full charge to 660 mAh in only 25.3 minutes,

[0057] - Fig.15 This example shows the change of K value based on Log10 scale.

[0058] - Fig.16 Example of NLV charging: the relationship between "K value" and "SOC" and time,

[0059] - Fig.17 The relationship between "K value" and SOC is illustrated.

[0060] - Fig.18 Shows the results of random test 3, where 13450 battery cells: 98% charge increase in only 22 minutes,

[0061] - Fig.19 Indicates the K value voltage ramp compensation when the current decreases.

[0062] - Fig. 20 The charge capacity and discharge capacity versus cycle number during NLV-based charging are illustrated with an average charge time of 23.45 minutes.

[0063] - Fig.21 illustrates the relationship between current and time during multiple charging cycles using NLV,

[0064] - Fig. 22 illustrates the voltage versus time relationship over multiple NLV-based charging cycles,

[0065] - Fig.23 The NLV charging curve measured for a 2100mAh battery cell is shown as an example.

[0066] - Fig.24 represents the temperature analysis of LIR 13450 cylindrical cells,

[0067] - Fig.25 exemplifies the charge / discharge curve of the NLV only with a duration of 36 minutes (average value),

[0068] - Fig.26 Indicates a daily charging plan with 30-minute fast charging,

[0069] - Fig. 27 The charge / discharge capacity or battery cell with a 36 minute charge time is illustrated, and

[0070] - Fig.28 Comparison of NLV and CCCV is characterized with a charging time of 30 minutes. DETAILED DESCRIPTION

[0071] This adaptive charging protocol (ACP) is based on non-linear voltammetry (NLV) based control during the period of charging the battery. This ACP allows the battery to be charged at different voltage levels with acceptable current (Amps) amounts based on its own state of health (SOH) and state of charge (SOC). Therefore, this protocol does not control or impose the amount of current drawn into the battery at any time.

[0072] Even though it is predicted that a battery can be charged (more than 80%) in less than 25 minutes using this method, the time may be longer or shorter based on the state of health (SOH) of the battery while charging. This method also ensures better safety compared to other fast charging methods [2, 3, 5], which mostly impose high current (I) in different modes / waveforms. Therefore, most importantly, this ACP method does not damage the battery by drawing a large fixed load of electrons through the battery cells without considering the health of the battery.

[0073] The kinetic balance of cell-particle dynamics (e.g., lithiation / delithiation (intercalation / deintercalation), ejection / floating of ions / electrons through solvents and separators, transport of charge against internal impedance (IR), etc. [4, 6]) determines how healthy the battery is? How much current can the battery system absorb / provide at one time during charge / discharge? We believe that this balance can be expressed as the “current change rate ” and “voltage change rate ". Therefore, the following relationship is used in forming this agreement:

[0074]

[0075] in:

[0076] K n : The "K value" is a constant during a certain period of time during charging.

[0077] n>=0

[0078] This is the rate of change of voltage (V) during the charging process [(Vstepend - Vstepstart) / steptime duration].

[0079] This is the absolute value of the rate of change of current (I) during the charging process.

[0080] α: This is an adjustable coefficient 0 < α < 100 and this coefficient gives the equation room to find a non-linear relationship between current and voltage based on different types of batteries. The system will also train it to best suit the charging process.

[0081] In addition, the relationship for α=1 can be simplified to:

[0082]

[0083] From the literature on lithium-ion batteries, it is evident that the battery chemistry provides inherent characteristics on the voltage curve. Within certain lower voltages (low SOC), the battery cell attempts to draw very low currents (due to high impedance), while at higher voltages (high SOC where polarization is lower), the potential to draw high currents is very high [1]. The voltage range within which some battery cells can tolerate these high currents is very narrow. Therefore, fast charging should be applied to keep the battery within these ranges for as long as possible until the expected capacity is obtained during the charging process (as much capacity as possible before the tolerable current drops below some lower level that will extend the total charging time).

[0084] All examples given below are related to lithium-ion batteries. However, ACP is applicable to all types of rechargeable batteries, including but not limited to solid-state lithium, NiMH, NiCd, LAB, alkaline cells, NaS, NaNiCl2, redox flow (ZnBr, VRB), ...

[0085] "ACP on NLV" refers to an Adaptive Charging Protocol (ACP) based on Non-Linear Voltammetry (NLV) charging. The protocol is adaptive because it adjusts to multiple user-driven factors and system / battery-driven factors to adjust its own charging curve to better respond to a given charging requirement. The user's expected charging time (duration), expected charge percentage (100%, 80% or 60%, etc.), possible relaxation time and initial state of charge (SOC) are some of the user-driven factors that adjust the process. Identifying the current SOC can also be designed to be processed automatically using entropy and enthalpy based methods (which are also system / battery driven factors). The state of health (SOH), specified (nominal) capacity, safe voltage range, available voltage control accuracy and the polarization curve of the battery are some of the factors in the automatic detection / system driven factors.

[0086] During NLV charging, the battery cell is set to a certain voltage (CV) that changes nonlinearly and increases gradually at each step. Thus, through a series of fast charging steps, the battery is charged for a period based on the nonlinear voltage (NLV).

[0087] During each of these steps, the cell draws a certain amount of current based on both its state of charge (SOC) and state of health (SOH) at very specific times. The current will then gradually drop. How fast the current drops at a certain step provides some clues about how well or poorly the battery wants to hold up in that NLV step. This way, more step time can be allocated whenever the battery is eager to draw more current, and less step time can be allocated when the battery is trying to dramatically reduce its current draw.

[0088] After each step, a very short relaxation with zero (0) current is applied to the system to stabilize the system with its new charge so that the OCV will drop to its stable (or almost stable) level. This creates a better chance for the next NLV charging step to obtain the optimal current based on its state without applying a high current beforehand [7, 8]. In this way, the protocol trains the battery cells to be stable and healthy (relaxing as much as possible without wasting too much time in too long relaxations) after each step and provides them with better preparation for the next step, allowing them to obtain more current than without relaxation. However, if the current drop in a certain step is not significant, the system is allowed to stay in that step for a longer time without moving to the next step. In this case, the rate of current drop and the maximum allowed time for such continuation of the step are monitored to decide when to move to the next step.

[0089] The system decides the "maximum allowed time for such continuation of the step" based on the adjustment parameters. Therefore, whenever a rapid drop in current is detected or the "maximum allowed time for the step" is exceeded, the system moves to the next charging step. Therefore, the actual time taken for a full charge depends on both the SOC and SOH of the battery.

[0090] In addition, the charging system uses three parameters to determine charge termination. First, whether the battery is fully charged based on the specified and obtained capacity. The second parameter is whether the maximum target termination voltage is reached. The system automatically adjusts the "target termination voltage" based on the polarization data of the relevant battery type / chemistry. The third factor (optional) is a self-learning model of the charge curve used to determine the charge state based on the real-time parameters while charging (by checking within a certain time window).

[0091] Figure 1 The current and voltage curves during the NLV charging process are shown. Some steps stay at the same voltage for longer than others. Generally, these steps can draw more current without a sharp drop over time. Therefore, the system is more stable and has the potential to accept more current and transfer more charge within the battery cell. It is also clear that at the beginning, the individual voltage steps are given very short durations and change rapidly until they start drawing some high current. This is also a good example of fighting high polarization at lower SOC [9, 10].

[0092] Additionally, frequent relaxations have been applied during this period.Finally, a similar situation can also be seen in the case of frequently changing steps with multiple relaxations, when the high current drawn is unstable and tends to drop rapidly.

[0093] Figure 2 The NLV-based charging process is shown as a flowchart of important processes. The following section will explain the details of each of these main processes.

[0094] [A] Initial SOC Discovery

[0095] This is an optional process as the system depends on SOC gain. Using any external method to do this measurement will help the system improve its performance.

[0096] Therefore, multiple methods have been explored to determine the initial SOC. Thermodynamic-based SOC prediction using fuzzy logic is one of the methods that have been identified as accurate and fast. Some other possible methods can also be found in references [11, 12]. Therefore, the system not only meets the charging from 0% (SOC) to 100% (SOC), but also supports any partial charging. Figure 3, this initial SOC (if available) can also be used to determine the initial “K” value.

[0097] [B] Initialize ACP-NLV charging

[0098] The initialization parameters of this protocol can be categorized into two main parts:

[0099] 1) User preference parameters

[0100] a) Expected Duration of Charge (ETD): Will charging be done within 15 minutes, 30 minutes, etc.

[0101] b) Enable Constant Current [CC] Leverage: Selecting this will force the system to use CC based charging before applying NLV based charging. However this is optional. The default current for this CC charging is 3C and the duration is 3 minutes. However these are configurable.

[0102] 2) System parameters

[0103] a) Start Voltage (startVoltage): The OCV of the battery just before charging begins.

[0104] b) nlvStatedCapacity (SC): The manufacturer-defined / experimentally proven battery capacity. This will be used to calculate the C-rate.

[0105] c) nlvStepInterval (CST): The duration of one step of charging / nlv changing. NLV charging is modeled as a series of shorter steps. During each step, the next expected voltage is calculated, charged and related parameters are updated.

[0106] d) nlv steps per frame (CFS:nlvStepPerFrame): the number of steps considered as a frame.

[0107] A frame is a collection of adjacent steps that are bundled together to apply some mathematical formula (average derivative of I and V) to determine the next possible voltage. The entire frame will first be filled with a series of known LV-based charging steps equal to the number of steps within the frame (CFS). Subsequently, after each step (CTS), only the last element of the frame will be updated with new data, while all other elements will be shifted back one position, discarding the first element. This will be performed during the entire charging process.

[0108] e) Frame current array (frameCurrentArr): Frame array of current that will get updated after each step time (CTS) duration.

[0109] f) Frame Voltage Array (frameVoltageArr): A frame array of current that will get updated after each step time (CTS) duration.

[0110] g) nlvInitialSlopeDuration: This is the time taken for the linear voltammetry (LV) based charging.

[0111] h) lvEquation: The linear equation used to generate the LV value in the very first stage of NLV charging [filling the first frame].

[0112] i) nlvEndTargetVoltage: The end voltage calculated based on the charge of LV. By default this is 4.2V.

[0113] j) nlv initial voltage array (nlvInitVoltArr): voltage array with frame size (CFS) based on the voltage step of LV.

[0114] k) nlv K value (nlvKValue): The first / default Kn value to start with [assumed to be predefined by LV-based analysis on the battery]. Later, the system will train this K value to support faster charging.

[0115] l) Minimum voltage (minVoltage) and maximum voltage (maxVoltage) (Vmax, Vmin): The minimum voltage / maximum voltage supported by the system for safety purposes.

[0116] m) Real-time C-rate (cRateRealTime): Based on the current drawn, the real-time C-rate will be measured: For example, C-rate = current / nlv specified capacity

[0117] n) Minimum expected C-rate (cRateMinExpected): The minimum required C-rate expected if the system is to be fully charged within the ETD time defined above.

[0118] o) C-Rate Drop Threshold (cRateDropThreshold): This is the "significant current drop" allowed in a single voltage step. If the C-Rate drops below this threshold, the system moves to the next step.

[0119] p)nlv Elapsed Charge Time (nlvElaspsedChargeTime): [nlvECT] is the duration spent charging at any given time.

[0120] [C] Apply CC [constant current] charging for 3 minutes

[0121] If this mode is selected, a 3C constant current (CC) will be applied for a short period of time to allow the battery (leverage) to charge quickly. The default period is 3 minutes, but both the CC current and this short period are configurable.

[0122] In CC charging, relaxation was applied after every 10th step [Cl REST, "0" current for 1 step time (CTS)]. Once CC-based charging was completed, a longer relaxation (3CTS) was applied before moving to the next treatment.

[0123] [D] Initial frame, LV-based charging

[0124] This step serves as an initialization / kick start process for NLV charging. A frame of current and voltage values ​​is needed for the NLV process to calculate the next non-linear setting voltage. Therefore, as a starting point, some other method is needed to charge the battery in a very short period of time (1 frame duration). This will also gain some capacity, which will keep the battery away from lower SOC levels where high polarization would prevent fast charging.

[0125] Therefore, any of the following methods are suitable for this launch:

[0126] -Charge based on linear sweep voltammetry (LSV)

[0127] -Charging based on Linear Sweep Amperemetry (LSA)

[0128] -Charging based on constant current constant voltage (CCCV)

[0129] - Cascaded Pulse Charging Protocol (CPC) described in PCT Application #PCT / IB2018 / 059705

[0130] To simplify the explanation, LSV has been used as the starting method:

[0131] 1. After the above relaxation, read the starting voltage (Vstart) of the battery.

[0132] 2. Read the expected initial termination voltage (Vinit-end) of the battery.

[0133] This is set up to achieve a greater high voltage range without compromising the battery.

[0134] 3. Use linear voltammetry (LSV) and acquire the array of voltage elements [array-lv] to start the charging process.

[0135] Vnext=Vstart+time interval*Vslope

[0136] Vslope is taken to charge the battery in 20 minutes [if ETD = 20 minutes]

[0137] 4. Define the following configuration parameters:

[0138] a. Measurement interval (ts)

[0139] b. Step interval (dt) [default is CTS]

[0140] c. Number of steps per frame (frame size, CFS)

[0141] d. Specified capacity (specified capacity)

[0142] e. Maximum allowed voltage (Vmax-end)

[0143] f. "Frame Size" VoltageFrame array size

[0144] g. "Frame Size" CurrentFrame array size

[0145] 5. The step of taking a frame from the LV array.

[0146] As defined in step 3, select a sub-array of "frame-size" elements from "array-lv".

[0147] 6. Start applying linear voltammetry based charging by following a voltage sequence based on "array-lv" for a duration of just 1 frame. The current and voltage are measured at various steps and populated into the relevant frame arrays. These will be used to calculate the slopes of the current and voltage changes.

[0148] [E] Update data frame (V, I, T) and capacity

[0149] Each step is followed by an update of the voltage (V), current (I) and temperature (T). Therefore, for each step, the update is done before triggering the next step. Therefore, the current taken to calculate the capacity increase is the minimum current during that CTS time frame (2 seconds by default). Further:

[0150] Each incoming / new reading will be stored as the next element in the frame according to "Update Path X". Since "Process D" will continue only CFS times, the frame will be completely filled when "Process D" is completed.

[0151] According to "Update Path Y", each new / next reading will be stored as the last element of the frame. All its previous data will be moved back by 1 position. Therefore, the exact first item of the frame will be erased each time.

[0152] The VoltageFrame and CurrentFrame arrays will be populated to store the frame values ​​and will be continuously updated during the charging process.

[0153] Update capacity:

[0154] An easy way to calculate SOC is to use real-time coulomb counting:

[0155]

[0156] in,

[0157] Reference Figure 5 , "I" is the current, and "dt" is the duration of a step, and Q nom is the nominal capacity of the battery cell.

[0158] The default "step time" has been set to 2s. Therefore, every time the battery draws some current, the associated capacity increase is calculated based on the above equation (C = I x t: current x time). This is then updated to the main capacity increase. This will be used in the protocol to define the SOC and subsequently control the parameters for changing the SOC.

[0159] There is no capacity calculation during the relaxation step.

[0160] [F] Find the next "set voltage" based on NLV [calculate derivative]

[0161] 1. Perform the following calculations using the VoltageFrame and CurrentFrame arrays and define the Next NLV-based Voltage (V-nlv-next).

[0162] a. Calculate the average value of the current derivative of the running frame (d[I] / dt(mA / s))

[0163] b. Get the last voltage from "VoltageFrame" ("V-nlv-prev")

[0164] c. Limit the “K value”

[0165] i. For the very first time, the K value will be a predefined value based on LV analysis of this type of battery at a 20 minute (ETD) LV charge.

[0166] ii. For subsequent processing, the system will train the "K value" based on a set of parameters (such as the C-rate drawn, the current drop rate within a step, the SOC, etc. This will be described in the section "Managing Self-Trained "K"").

[0167] d. Define V-nlv-next as follows:

[0168] V-nlv-next=V-nlv-prev+K value*[{1 / (d[I] / dt(mA / s))}*dt]

[0169] e. Check the previous NLV step to identify whether there is a current drop or increase, and whether there is a drop (Idrop = [(lx + 1 - Ix) / Ix)]), then apply voltage compensation to reduce the next voltage ramp rate.

[0170] V-nlv-next=V-nlv-next-Idrop*(V-nlv-next-V-nlv-prev)

[0171] 2.T, temperature will also be used to ensure the safety of the battery during charging. Whenever the temperature rises above the safety limit, the charge will be held and let the system relax for a predefined amount of duration until it is restored when the expected temperature range is ensured.

[0172] [G] Using NLV to charge

[0173] 1. Set "V-nlv-next" to battery

[0174] a. After this setting, immediately read the current (realtimeCurrent) and voltage (realtimeVoltage).

[0175] b. Read the current and voltage after the "step interval"

[0176] c. Take the average of the above two items and update the last item of VoltageFrame and CurrentFrame.

[0177] 2. Read the current drawn immediately after setting V and compare to safety limits to avoid any damage in case of any unexpectedly high current. Otherwise wait for the CTS duration until the next "NLV Set Voltage" is set. There is no imposition on current. The system simply measures current and voltage at its best possible frequency.

[0178] [H] Managing the "K" and [I] step times of self-training

[0179]

[0180] The above equation is used to determine the NLV-based setting voltage for each individual charging step. However, K n It also varies based on a set of factors. The following are the main factors used to control it:

[0181] The expected C-rate: cRateExpected is used to ensure that the required capacity is achieved within the required duration at full charge.

[0182] Based on the user's preferences / requirements for "Charge Time" and "Charge Capacity", the system can calculate a minimum C-rate ("Expected C-rate") that must be maintained continuously or as an average value during the entire charge period. The protocol uses this information to control Kn and step time by comparing said minimum C-rate ("Expected C-rate") with the C-rate driven by the real-time current in each charging step ("Real-time C-rate: cRateRealTime(CRRT)").

[0183] Whenever a high "Real-time C-rate" is drawn, Kn is kept as low as possible. And the step time is increased as much as possible. At the same time, without applying relaxation, the "step time" will not be allowed to exceed the "maximum allowed time for a step". However, if the system draws a high "Real-time C-rate" even after relaxation, it allows the same "voltage step" to continue until a "significant current drop" is identified (this is a system configurable parameter). It will then decide to move to the next voltage step.

[0184] The "elapsed charge time: timeElapsedCharge" is used to ensure that the required charge is achieved within the expected duration.

[0185] This will also be factored into the state of charge (SOC). When the end of the expected charge duration is reached, the system will increase the charge frequency by reducing the step time and increasing Kn to a higher value to quickly scan the non-linear voltage change.

[0186] However, if the current drawn by the system is close to or within the range of the "expected C-rate", the system will maintain the nominal range of step times and Kn values.

[0187] The "C-rate consumption duration: timeWaitedForExpectedCRate" is used to try and push the system out of such a high-resistant charging window.

[0188] Whenever the system detects that the current drawn at a voltage step is well below the "expected C-rate" threshold, it will try to go through the step as quickly as possible. Thus, the "step time" will be reduced.

[0189] However, if this happens at a very initial stage (at low SOC), the Kn value will increase greatly, causing a large increase in voltage.

[0190] If it occurs near the end of charging, the Kn value will remain at a moderate level, since there is still a lot of charging to be done to get more capacity. Here, the expected value of the "C-rate" may also drop to half of its full expected value.

[0191] In this case, when the "step time" is reduced, the system attempts to speed up the sweep of the charging steps. Therefore, in some cases, the current drawn may rise again. But in other cases, the current drawn may remain at a lower C-rate. In this lower case, the Kn value will be set to a very high value until the battery can draw a fairly acceptable current level. Whenever it starts to reverse (start-back) to draw a high C-rate current, the Kn value will be reduced, but the "step time" is kept small to pass this difficult period as quickly as possible, while obtaining the maximum possible charge even during this period.

[0192] Based on what is claimed above, the control logic and reference table we use for the reference protocol are as follows:

[0193]

[0194] Table 1.1 Reference table for different levels of "cRateExpected"

[0195]

[0196] Table 1.2 Reference table for different levels of "nlvKValue_TrainedFactor"

[0197]

[0198] Table 1.3 Reference table for different levels of "timeWaitedForExpectedCRate"

[0199]

[0200] Table 1.4 Reference table for different levels of "timeElapsedCharge"

[0201]

[0202] Table 1.5 Reference table for different levels of "steptimeFactor"

[0203] Figure 6 The process in illustrates how to control the K value and step time based on the parameters shown in the above tables [1.1 to 1.5].

[0204] according to Figure 7 , the "default K (K*k_TF)" and "default step time" will be used only for the very first step of NLV-based charging. For all subsequent steps, the above training algorithm will be applied and the most suitable "K value" and "step time" will be found.

[0205] Therefore, it is ensured that these parameters are adjusted based on the SOC and SOH of the battery, which makes the possible draw current different.

[0206] like Figure 7 As shown, whenever the C-rate drawn is very low, the K value increases rapidly. However, within the space that the battery can handle, the current tries to increase gradually.

[0207] Likewise, when the C-rate is high, the K value decreases. However, the K value only decreases to very low values ​​when the system attempts to draw current with a C-rate close to or higher than the expected C-rate.

[0208] like Figure 8 As shown, when the SOC is about half of full capacity, the K value becomes very low because the battery has a high potential to draw high current.

[0209] like Fig. 9 As shown, when the SOC approaches full capacity, the K value fluctuates rapidly and tries to maintain a higher draw current.

[0210] [J] Adjust the "Target Termination Voltage"

[0211] The idea of ​​having an adjustable "target termination voltage" is to enhance the capacity obtained according to its SOC and SOH. Whenever the battery has a good SOH, most of the charging capacity can be drawn within the preferred voltage range. Therefore, the system sets the "default target termination voltage" as the exit point of NLV charging at the beginning. Whenever the real-time voltage of the battery reaches this "default target termination voltage", the system checks the C-rate driven by the real-time current at that time. Then, based on this C-rate, the system determines whether to increase the "target termination voltage" or continue charging or stop charging at this point. In order to determine this based on the C-rate, two methods are considered in the protocol:

[0212] Acceptable "target termination voltage" based on a specific polarization curve

[0213] Here, the system needs pre-processing information about the polarization curve of the battery type, which is used as the target for charging. Therefore, the protocol should be tuned for different battery types.

[0214] Default Target Termination Voltage Table

[0215] The table is intended to be used as a universal controller of the termination voltage for any battery type. In addition, the table is formed as an average polarization curve that can be used whenever a "specific polarization curve" is not suitable for the battery type intended to be charged. It is also intended to train itself based on charging statistics.

[0216] The following Table 1.6 is used as a "default termination voltage table" for the reference protocol described herein:

[0217]

[0218] Table 1.6 Default termination voltage table based on empirical data

[0219] If the "Default Target Termination Voltage" is selected as 4.65V, then Table 1.6 corresponds to the termination voltage values. But again this is a customizable parameter that can be changed under system / user preference. However, we intend to have a range for this based on battery type / chemistry. Therefore, as a global control logic, the operation to handle the "Adjustable Termination Voltage" is as follows Fig.10 shown.

[0220] [K] Exit Criteria

[0221] There are three different criteria for deciding when to stop the charging process.

[0222] 1. Once the system reaches the "maximum termination voltage"

[0223] 2. Once the battery has reached the required full capacity

[0224] 3. Comparison of current curves based on learning

[0225] If the current profile closely matches any previous current profile seen during similar exit situations, the learning algorithm intends to improve its exit profile. Depending on the availability of the above three methods, the same priority as 1, 2, and 3 will be considered to decide whether to exit.

[0226] [L] Management REST

[0227] Management Rest always applies zero (0) current to the battery. The charging cycle will be spent during this Rest period.

[0228] [M] Exit NLV

[0229] Once at least one of the criteria is reached, NLV charging stops. However, depending on how much capacity increase is achieved, the system decides to do another CC [using 2C constant current charging] or NLV again.

[0230] [N] Apply termination CC

[0231] If the capacity of the NLV drive is insufficient compared to the target, a 2C constant current charge will be applied during the 2 minutes when the NLV charge is terminated to further gain capacity. This constant current and its duration can be configured as system parameters.

[0232] Reference Fig.11 and Fig.12 , this is the average curve of a test that went through 24 cycles without any problems until we stopped it purposefully.

[0233] Alternatively, according to the present invention, a CC protocol, a CCCV protocol and / or a cascade pulse charging protocol (PCT application #PCT / IB2018 / 059705) may be applied at the beginning, middle and end of an NLV protocol.

[0234] Multi-stage k value management

[0235] The K value is changed based on how best the battery can draw the expected C-rate above the current. If the battery draws a very low C-rate, the K value will increase rapidly to model the sudden rise in voltage and subsequent high current. If the battery draws the expected C-rate or higher, the K value is changed to a very low value and a best effort is made to get the maximum possible charge with this high current charge. In other cases, the K value is changed to maintain the expected C-rate as long as possible.

[0236] Fig.13 Illustrated is the NLV charging of Random Test 1: Current versus K value during NLV charging ending with a full charge to 660 mAh in only 23.3 minutes.

[0237] Fig.14 Illustrated is the NLV charging of Random Test 2: Current versus K value during NLV charging to fully charge to 660 mAh in 25.3 minutes.

[0238] Fig.15 The variation of K over time is shown in logarithmic scale.

[0239] For NLV charging, Fig.16 It shows the change of “K value” and SOC over time, and Fig.17 The evolution of the “K” value with SOC is shown.

[0240] Fig.18 The curve in shows the charging curve of the NLV-based adaptive charging protocol applied to a 13450 battery cell. This has charged a battery of a specified capacity of 700mAh to 99% in only 28 minutes, obtaining a capacity of 695mAh.

[0241] During this processing period,

[0242] - It obtained a charge based on 2C or more during half of the process.

[0243] -A quarter of the full charge time also drew over 3C, and this happened at the very end.

[0244] The highlighted segments were further analyzed to envision how they work in the protocol.

[0245] from Fig.19 The highlighted part analyzes about 100 samples:

[0246] -Selected K: This is the K that is pre-selected and added to the ACP

[0247] - Generated K: This is the K recalculated based on the derived "V-nlv-next" using the above K.

[0248] - Both K above are multiplied by 1000 to clearly show the changes in the figure below.

[0249] - As shown in the next section, this variation seen in the "generated K" vs. the "selected K" is due to precision errors, as the voltage rise is always a small amount (mV):

[0250] -A: The K generated during this period is the same as the selected K

[0251] - B: The K generated during this period deviates from the selected K.

[0252] The above "B" segment is shown in the rectangular box in the table below.

[0253] "AVG(Abs(dI / dt))" and "dV / dt" are calculated for each current and voltage change collected during the charging process.

[0254]

[0255] In addition, the formation of [dI / dt(n-1)] & [dI / dt(n-2)] can be achieved from the following table.

[0256]

[0257] As shown in the table above, the current drops during this "B" period.

[0258] Therefore, both dI and dv propagate sudden increases or decreases. This causes the multiplication accuracy to deviate in their product.

[0259] Comparison of charge capacity and available discharge capacity

[0260] Fig. 20 The discharge capacity is shown to remain almost constant at about 630 mAh, which is about 97% of the specified capacity of the battery. Thus, it is demonstrated that this NLV-based charging method does not contribute much to capacity decay over time.

[0261] This is a great advantage compared to other competing fast charging methods that are mainly based on directly applying high currents.

[0262] Fig.21 and Fig. 22 It is shown that even after multiple cycles of NLV based charging, the charging curves of current and voltage remain essentially the same. This is another demonstration showing that the battery cells are not damaged by this fast charging process.

[0263] Reference Figure 23 to Figure 28 , experimental tests of ACP charging have successfully completed 1,300 cycles with a charging time of 30 minutes. Fig.28 It is more specifically shown that ACP charging is safer than CCCV charging. In fact, after ten cycles, the battery charged by CCCV exploded, while the battery charged by NLVACP was preserved.

[0264] References

[0265] [1]SS Zhang, J. Power Sources 161(2006)1385-1391.

[0266] [2] Ronald Baroody, "Evaluation of rapid electric battery charging techniques", university of Nevada Las Wegas, 2009

[0267] [3]D.Ansean,V.M.Garcia,M.Gonzalez,“Efficient fast-charging strategiesfor Li-ion batteris”,University of Oviedo,Gijon,Spain.

[0268] [4]Venkatasailanathan Ramadesigan,Paul W.C.Northrop,Sumitava De,“Modeling and Simulation of Lithium-Ion Batteries from a Systems EngineeringPerspective”,Washington University,St.Louis,Missouri 63130,USA.

[0269] [5]Elie Ayoub,Nabil Karami,“Review on The Charging Techniques of aLi-ion Battery”,University of Balamand,Koura,Lebanon.

[0270] [6]Martin Z.Bazant,“Theory of Chemical Kinetics and Charge Transferbased on Nonequilibrium Thermodynamics”,Masachusetts Institute of Technology,Cambridge Masachusetts 02139,United States,2012.

[0271] [7]Liang He,Eugene Kim,Kang G.Shin,“*-Aware Charging of Lithium-ionBattery Cells”,The University of Michigan,Ann Arbor,MI,USA,2016.

[0272] [8]D.Andrea,“Battery management systems for large Lithium-ion batterypacks”,Artech House,2010.

[0273] [9]Jiang JC,Zhang CP,Wen JP,et al.An optimal charging method for Li-ion batteries using a fuzzy-control approach based on polarizationproperties.IEEE Trans Veh Technol2013;62(7):3000-9.

[0274]

[10] Jiang JC,Liu QJ,Zhang CP,et al.Evaluation of acceptable chargingcurrent of power Li-ion batteries based on polarization characteristics.IEEETrans Ind Electron 2014;61(12):6844-51.

[0275]

[11] I.-S.Kim,The novel state of charge estimation method for lithium-ion battery using sliding mode observer.Journal of Power Sources,163(1):584-590,2006.

[0276]

[12] X.Hu,F.Sun,Y.Zou,“Estimation of state of charge of a lithium-ionbattery pack for electric vehicles using an adaptive luenbergerobserver.Energies,2010.

Claims

1. A method for charging a battery system based on nonlinear voltammetry (NLV), the method comprising cycles for charging the battery system, each cycle for charging comprising the following steps: During the variable cycle duration, the battery system is set to a charging voltage that varies non-linearly and increases gradually at each cycle, Measure the current being drawn by the battery system to determine the time derivative of the charging current and monitors the rate of decrease of the current being drawn, measuring a temperature within the battery system, applying relaxation with at least zero current to the battery system with charge maintained whenever the measured temperature rises above a safety limit to stabilize with the new charge of the battery system until restored while ensuring the expected temperature range, and By applying the time derivative of the voltage calculated according to the following equation for the previous cycle duration To determine the next charging voltage value for the next charging cycle: in: K n According to a set of parameters including the state of charge SOC and the state of health SOH of the battery system, the α is an adjustable constant to match the nonlinear relationship between current and voltage based on different types of batteries, and is the absolute value of the charging current rate.

2. The method according to claim 1, wherein: The battery system includes one battery cell or a multi-battery cell system.

3. The method according to claim 2, wherein: The multi-battery cell system is arranged in a series and / or parallel battery cell configuration.

4. The method according to claim 2, wherein: The voltage of the battery cell is between 2V and 5V.

5. The method according to claim 2, wherein: The charging current in the battery cell is between 0C and 10C.

6. The method according to claim 2, wherein: The temperature T of the battery cell is between -20°C and +55°C.

7. The method according to claim 1, wherein: Charging time from 0% SOC to 100% SOC t ch Between 10 minutes and 2 hours.

8. The method according to claim 1, wherein: SOC ranges from 0% to 100%.

9. The method according to claim 1, wherein: The number of cycles is 200≤n≤2000.

10. The method according to claim 1, wherein: The NLV-based method is combined with a constant current CC protocol, a constant current constant voltage CCCV protocol, and a cascade pulse charging CPC protocol.

11. The method according to claim 10, wherein: Any one of the CC protocol, CCCV protocol, and CPC protocol is applied at the beginning, middle, and end of the NLV-based method.

12. A nonlinear voltammetry (NLV) based system for charging a battery system, the NLV based system implementing the method according to claim 1, the NLV based system comprising: means for setting the battery system to a charging voltage during a variable cycle duration, the charging voltage varying non-linearly and increasing gradually at each cycle, Used to measure the current being drawn by the battery system to determine the time derivative of the charging current and a device for monitoring the rate of decrease of the current being drawn, means for measuring the temperature within said battery system, means for applying relaxation with at least zero current to the battery system with charge maintained whenever the measured temperature rises above a safety limit to stabilize with the new charge of the battery system until recovery when the expected temperature range is ensured, and The time derivative of the voltage calculated according to the following equation is applied for the previous cycle duration Means for determining the next charging voltage value for the next charging cycle: in: K n According to a set of parameters including the state of charge SOC and the state of health SOH of the battery system, the α is an adjustable constant to match the nonlinear relationship between current and voltage based on different types of batteries, and is the absolute value of the charging current rate.

13. The system according to claim 12, wherein: The battery system includes one battery cell or a multi-battery cell system.

14. The system according to claim 13, wherein: The multi-battery cell system is arranged in a series and / or parallel battery cell configuration.

15. The system of claim 13, wherein: The voltage of the battery cell is between 2V and 5V.

16. The system of claim 13, wherein: The charging current in the battery cell is between 0C and 10C.

17. The system of claim 13, wherein: The temperature T of the battery cell is between -20°C and +55°C.

18. The system of claim 12, wherein: Charging time from 0% SOC to 100% SOC t ch Between 10 minutes and 2 hours.

19. The system of claim 12, wherein: SOC ranges from 0% to 100%.

20. The system of claim 12, wherein: The number of cycles is 200≤n≤2000.

Images