An active energy balancing method for source-load separation between battery groups

Abstract

The present invention relates to an active energy balancing method with source-load separation between battery groups, and belongs to the field of power electronics. In the prior art, most balancing topologies and balancing control methods can only balance adjacent batteries, and there are problems such as fixed balancing current, slow balancing speed, and low efficiency. The method described in the present invention connects the battery groups with the proposed balancing topology structure, and uses a source-load separation active energy balancing method based on fuzzy control and optimized droop control to improve the effect of power balancing. When the SOC difference between battery groups reaches a set threshold, the method described in the present invention can automatically determine the balancing mode and balancing parameters such as balancing current, so that energy can be balanced between any two battery groups, and the size of the balancing current can be determined according to the battery group status, thereby controlling the balancing speed, which greatly improves the flexibility, balancing speed and efficiency of the battery group SOC balancing.

Description

Technical Field

[0001] The invention belongs to the technical field of power electronics, and in particular relates to an active energy balancing method for source-load separation between battery groups. Background Art

[0002] With the rapid development of society, the demand for energy in social life and various industrial fields is also increasing. As a portable, clean and pollution-free energy source, batteries have been widely used in smart grid energy storage systems, DC microgrids, electric vehicles, new energy drones and other fields. A single battery has the problem of low voltage and small capacity, and multiple single batteries need to be used in series or parallel to meet different needs. However, in actual use, due to the differences in the batteries themselves and various external factors, the battery pack has the problem of inconsistent state of charge (SOC), which in turn affects the performance of the energy system. In other energy systems, battery packs are required to maintain similar power or perform specific energy exchange. In the aforementioned applications of battery packs, such as the power supply system of new energy drones, when multiple battery packs are used, if a battery pack is out of work due to SOC imbalance, it will have a serious impact on the load carried by the drone and its normal operation. Dynamic real-time balancing of SOC between battery packs plays a vital role in maintaining the stable operation of the entire energy system. Therefore, it is very necessary to improve the flexibility, stability, balancing speed and efficiency of the battery pack SOC balancing control strategy.

[0003] Yan Ma et al. proposed a two-level balancing topology within and between groups, and obtained the balancing current between two adjacent modules or cells through fuzzy control [Yan Ma et al. Equalization of lithium-ion battery pack based on fuzzy logic control in electric vehicle. IEEE Trans. Ind Electron., Aug. 2018, 65 (8): 6762-6771]. However, this method can only exchange energy between two adjacent modules or cells to achieve overall energy balance, and cannot make energy directly interact between any two battery groups. In some cases, there is a problem of low balancing efficiency. Xiaonan Lu et al. used an adaptive droop control method based on SOC to perform energy interaction and power distribution in distributed energy storage systems [Xiaonan Lu et al. State-of-charge balance using adaptive droop control for distributed energy storage systems in DC microgrid applications. IEEE Trans. Ind Electron., Jun. 2014, 61 (6): 2804-2815]. Although this method can dynamically adjust the droop coefficient according to the SOC situation, the droop coefficient is inversely proportional to the nth order of SOC. It is necessary to select a suitable n to achieve a more ideal effect, and in some cases, it is easy to produce an unreasonable droop coefficient. Thomas Morstyn et al. used a multi-agent sliding mode control to balance the charge state between distributed DC microgrid battery energy storage systems [Thomas Morstyn et al. Multi-agent sliding mode control for state of charge balancing between battery energy storage systems distributed in a DC microgrid. IEEE Transactions on Smart Grid., Sept. 2018, 9 (5): 4735-4743]. However, the characteristic of this method is that the battery energy storage system is either fully charged or fully discharged, which has limitations.Khanh Duc Hoang et al. proposed the concept of virtual rated power, which realizes accurate power distribution of distributed battery energy units in DC microgrid according to SOC level and virtual rated power, so as to achieve SOC balance of battery pack [Khanh Duc Hoang et al. Accurate power sharing with balanced battery state of charge in distributed DC microgrid. IEEE Trans. Ind Electron., Mar. 2019, 66 (3): 1883-1893]. However, it does not focus on direct energy exchange between battery packs to achieve energy balance. In addition, many balancing topologies and methods based on bidirectional DC converters (DC-DC bidirectional converters) have also been proposed.

[0004] Although the aforementioned methods have been verified through simulation or experiments, they still have a series of disadvantages, such as energy can only be exchanged between two adjacent modules or cells, it is difficult to adjust the balancing parameters, it cannot be directly balanced in the island mode, the balancing mode is single, the balancing efficiency is low, and the practicality of the method is not considered. Therefore, the existing energy balancing control methods between battery packs are rarely used in practice, which means that it is necessary to study more flexible and efficient energy balancing control methods between battery packs. Summary of the invention

[0005] In view of the defects existing in the prior art, the purpose of the present invention is to propose an active energy balancing method with source-load separation between battery groups. When there are certain differences in the SOC between battery groups, the balancing mode, balancing current and other parameters can be automatically determined, and the energy balancing speed between any battery groups can be dynamically controlled, thereby improving the flexibility, balancing speed and efficiency of the battery group SOC balancing.

[0006] To achieve the above purpose, the technical solution adopted by the present invention is: an active energy balancing method with source-load separation between battery packs. On the basis of connecting battery packs using a DC-DC bidirectional converter balancing topology hardware platform, active energy balancing between battery packs is performed; after performing source-load separation and balancing mode judgment according to the characteristics of the balancing topology, fuzzy control is used to obtain the load side (energy absorption side) balancing current and the source side (energy output side) droop coefficient, and then energy balancing is performed in combination with optimized droop control, including the following steps:

[0007] (1) Estimate the power of each phase battery group and use the source-load separation method to divide each phase battery group into the source side (energy output side) and the load side (energy absorption side). The corresponding converters are the source side converter and the load side converter respectively;

[0008] (2) Use a multi-value balancing strategy to determine whether each phase battery pack is balanced;

[0009] (3) Fuzzy control combined with optimized droop control method is used. After balancing is turned on, fuzzy control is used to calculate the balanced current I on each load side. equ_i Then determine the total load current on the bus and use fuzzy control to calculate the source side droop coefficient Droop i , combined with optimized droop control, power is distributed to each source side, so that energy is transmitted from the source side to the load side for energy balance;

[0010] (4) During the battery balancing process, the balancing mode and load-side balancing current I are dynamically updated according to the battery status. equ_i And the source side droop coefficient Droop i .

[0011] Furthermore, the DC-DC bidirectional converter balanced topology is a high-voltage side common DC bus structure, and any number of battery packs are respectively connected to the same number of low-voltage sides of each converter, and the high-voltage sides of each converter are connected in parallel to establish an internal balanced bus with no other loads on it.

[0012] Furthermore, the source-load separation method is: using the SOC of each phase battery pack i , calculate the average SOC of all battery packs avg , Maximum SOC max , minimum SOC min , range ΔSOC max , SOC of each phase battery pack i With SOC avg The difference ΔSOC i , according to ΔSOC i The positive or negative value of ΔSOC determines whether each phase battery pack is controlled as the source side or the load side. i is a positive value, the battery pack of this phase is controlled as the source side; ΔSOC i If it is a negative value, the battery pack of this phase is controlled as the load side.

[0013] Furthermore, the multi-value balancing strategy is: i Set different thresholds for opening and closing balancing. When the threshold conditions are met, the battery pack of this phase will open or close the balancing. max Set the threshold for turning on and off balancing. If the threshold condition is met, balancing is turned on or off for the battery group with the largest and smallest power.

[0014] Furthermore, the fuzzy control combined with the optimized droop control method is: after starting the balance, on the load side, the SOC i and ΔSOC i As the input of the load-side fuzzy controller, the load-side balanced current I is obtained according to the load-side fuzzy rules. equ_i ; On the source side, SOCi and ΔSOC i As the input of the source side fuzzy controller, the source side droop coefficient Droopi is obtained according to the formulated source side fuzzy rule. In the source side droop control link, a balanced internal bus voltage correction link is added. The input of this link is the difference between the bus voltage set value and the actual value. Proportional integral control is adopted, and the output is the bus voltage correction value, and the bus voltage set value can be changed.

[0015] Furthermore, during the balancing process, the battery pack status is updated in real time, and the balancing mode and load-side balancing current I are re-determined based on the battery status every once in a while. equ_i Calculate and update the source side droop coefficient Droopi.

[0016] The effects of the present invention are as follows: by adopting the method described in the present invention, energy balance between any two battery groups can be achieved, and the flexibility of balance is greatly enhanced; by dynamically adjusting the load-side balanced charging current and the source-side droop coefficient by fuzzy control, reasonable power distribution is performed, that is, the balancing current is controllable, and overcharging and over-discharging of the battery are avoided; the balancing speed and energy transfer efficiency are greatly improved; the stability of the internal DC bus is maintained through the bus correction link, and the bus voltage fluctuation caused by the droop control is eliminated; in summary, by adopting the method described in the present invention, when the SOC difference between the battery groups reaches the set threshold, the balancing mode and the balancing current and other parameters can be automatically determined, and the energy balancing speed between any battery groups can be dynamically controlled, which greatly improves the flexibility, balancing speed and efficiency of SOC balancing between battery groups. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is the overall control flow chart of active energy balancing with source-load separation between battery packs;

[0018] Figure 2 It is a block diagram of source-load separation active energy balancing control;

[0019] Figure 3 This is a topological diagram of an inter-group balancing circuit based on a DC-DC bidirectional converter;

[0020] Figure 4 It is the block diagram of fuzzy logic control for balanced load side between groups;

[0021] Figure 5 is the membership function of the load-side input and output;

[0022] Figure 6 is the fuzzy rule table on the load side;

[0023] Figure 7 It is the block diagram of fuzzy logic control on the source side of inter-group balance;

[0024] Figure 8 is the membership function of the input and output quantity on the source side;

[0025] Fig. 9 is the source side fuzzy rule table;

[0026] Fig.10 It is the setting change diagram of SOC state;

[0027] Fig.11 is a graph of energy balance parameter changes using this method;

[0028] Fig.12 This is the result of the energy balance experiment between groups using this method. DETAILED DESCRIPTION

[0029] The specific implementation modes of the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0030] The present invention is designed to provide an active energy balancing method with source-load separation between battery packs. On the basis of connecting the batteries with the proposed balancing topology structure, source-load separation and mode judgment are performed through more reasonable balancing control. Fuzzy control and optimized droop control are used to determine the balancing mode parameters according to the SOC changes of the battery pack, and a bus compensation link is added to improve the flexibility, balancing speed and efficiency of the SOC balancing of the battery pack. Figure 1 is the overall control flow chart of the present invention, Figure 2 It is a control block diagram of the present invention.

[0031] The specific implementation steps are as follows:

[0032] Step 1: Connect four 48V battery packs to the low-voltage side of the Buck / Boost bidirectional converter according to the balanced topology, and connect the high-voltage side of the converter in parallel to the common internal balanced DC bus. Figure 3 The low-voltage side voltage U of each phase is obtained by hardware sampling of each phase. Low_i , low voltage side current I Low_i , high voltage side current I High_i And the common internal bus voltage U dc .

[0033] The voltage and current data obtained by hardware circuit sampling are used to estimate the SOC of each phase battery pack based on the OCV-AH method. Important parameters are calculated and used as a basis for source-load separation. The specific implementation details are as follows.

[0034] Before the system starts balancing, obtain the low-voltage side battery pack voltage U Low_i , which can be regarded as the battery open circuit voltage before the system is running, and the initial value SOC of each phase battery pack is determined by the battery OCV-SOC curve 0_i .

[0035] After obtaining the initial value of SOC, real-time ampere-hour integration is performed, but the traditional ampere-hour integration method does not consider the influence of factors such as temperature and charge-discharge efficiency on the remaining power estimation, which makes the accuracy of SOC estimation not high. The present invention has made certain optimizations and compensated for the temperature and charge-discharge efficiency coefficient in the calculation, which can further improve the estimation accuracy of SOC.

[0036] The battery's charge and discharge conditions are different at different temperatures, so the impact of temperature on SOC estimation cannot be ignored. The commonly used temperature compensation formula is as follows.

[0037] η T =1+0.008(T N -T) (1)

[0038] Among them, T N refers to the standard temperature of 20℃, and T refers to the current temperature value.

[0039] The current of the battery will change greatly under different working conditions, which will affect the charging and discharging efficiency of the battery. If the influence of the charging and discharging efficiency of the battery is not considered, the cumulative error of SOC will become larger and larger, resulting in inaccurate estimation results. The commonly used relationship between available power and discharge current is as follows.

[0040] Q=A*I n-1 (2)

[0041] Among them, I represents the discharge current, n refers to the structural constant of the battery (generally 1.15-1.42), here it is 1.2, and A is the battery constant related to the active material. Under the same initial conditions, the values ​​of A and n are the same, also 1.2, so the charge and discharge efficiency can be obtained as follows.

[0042]

[0043] Among them, I N is the rated current, Q N is the rated capacity.

[0044] Through the above analysis, the corrected SOC mathematical model is established as follows:

[0045]

[0046] Among them, SOC 0_i Refers to the initial value of the charge of the i-th battery pack before discharge; η T Represents the temperature compensation coefficient; η i I represents the charging and discharging efficiency compensation coefficient of the i-th battery pack; Low_i (t) refers to the charging and discharging current of the ith battery pack, which is a function of time t; Q 0_iis the capacity of the ith battery pack before discharge.

[0047] Calculate the average, maximum, minimum, range, and SOC of each phase i The difference from the average value (mean deviation value) is as follows.

[0048]

[0049] SOC max =MAX{SOC1,SOC2,…SOC i …,SOC N} (6)

[0050] SOC min =MIN{SOC1,SOC2,…SOC i …,SOC N} (7)

[0051] ΔSOC max =SOC max -SOC min (8)

[0052] ΔSOC i =SOC i -SOC avg (9)

[0053] Where N refers to the number of battery packs. The power of each battery pack is dynamically ranked, and each battery pack is divided into a source side and a load side using the source-load separation method. The detailed process is as follows.

[0054] The condition for judging the i-th phase battery group as the "load" side is

[0055] ΔSOC i ≤0 (10)

[0056] The condition for determining the i-th phase battery group as the "source" side is

[0057] ΔSOC i >0 (11)

[0058] Step 2: Use the multi-value balancing strategy of "average deviation value + extreme value + multiple thresholds" to calculate the average deviation value based on ΔSOC. i and ΔSOC max Determine whether each phase is balanced. The detailed process is as follows.

[0059] After the load side and the source side are separated, the condition for starting the balancing of the i-th phase battery group is

[0060] |ΔSOC i |≥λ1 (12)

[0061] Among them, λ1 is the threshold for starting equalization, which is 3 here.

[0062] The condition for closing the balancing of the i-th phase battery group is

[0063] |ΔSOC i |≤λ2 (13)

[0064] Among them, λ2 is the closing balance threshold, which is 1 here.

[0065] In addition, the balancing problem of the battery with the largest power and the battery with the smallest power is considered. That is, the condition for starting balancing based on the range is:

[0066] |ΔSOC max |≥λ3 (14)

[0067] The condition for closing equilibrium by range is

[0068] |ΔSOC max |≤λ3 (15)

[0069] Among them, λ3 is the threshold for turning on and off balancing for extreme difference judgment, which is 4 here. If the above conditions are met, the battery with the largest power and the battery with the smallest power will turn on and off balancing accordingly.

[0070] Step three, use fuzzy control combined with optimized droop control method to balance energy. The specific implementation details are as follows.

[0071] For the load side, after starting the balance, SOC i and ΔSOC i As the input of the load-side fuzzy controller, the output is the phase balanced current I equ_i ,like Figure 4 The load-side fuzzy controller converts the precise input into fuzzy variables, which are processed by the established load-side fuzzy rule base. The result of the inference engine is converted into precise output by the defuzzifier. The detailed implementation process is as follows.

[0072] Input SOC i and ΔSOC i The domains of are set to (0, 100%) and (0, 70%) respectively. The converted fuzzy variables are divided into five fuzzy subsets: very small (VS), small (S), medium (M), large (L), and very large (VL). The output I equ_i The domain of is set to (0,3) according to the reasonable current range, and the fuzzy variables corresponding to the interval are also divided into five sub-intervals: very small (VS), small (S), medium (M), large (L), and very large (VL).

[0073] The membership function selects triangular fuzzy membership function, and the load side input and output membership function is as follows: Figure 5 shown.

[0074] According to the battery charging and discharging data and experience, a fuzzy rule table for the load side is established, such as Figure 6 shown.

[0075] The centroid defuzzification method is used to obtain the balanced current I equ_i , whose expression is

[0076]

[0077] The balancing current I equ_i As a given current, the load side is controlled by a single current loop through a bidirectional DC-DC converter and is controlled into a current source mode.

[0078] For the source side, after starting the balancing, the SOC i and ΔSOC i As the input of the source side fuzzy controller, the output is the phase droop coefficient Droop i ,like Figure 7 The source-side fuzzy controller converts precise input into fuzzy variables, which are processed using the established source-side fuzzy rule base. The result of the inference engine is converted into precise output by the defuzzifier. The detailed implementation process is as follows.

[0079] Input SOC i and ΔSOC i The domains of are set to (0, 100%) and (0, 70%) respectively. The converted fuzzy variables are divided into five fuzzy subsets: very small (VS), small (S), medium (M), large (L), and very large (VL). The output Droop i The domain of is set to (1.6,3) according to the reasonable parameter range of droop control, and the fuzzy variables corresponding to the interval are also divided into five sub-intervals: very small (VS), small (S), medium (M), large (L), and very large (VL).

[0080] The membership function selects the triangular fuzzy membership function, and the source side input and output membership functions are as follows: Figure 8 shown.

[0081] According to experience, a fuzzy rule table is established on the source side, such as Fig. 9 shown.

[0082] The centroid defuzzification method is used to obtain the balanced current Droop for the fuzzy output. i , whose expression is

[0083]

[0084] After obtaining the source side droop coefficient, add the bus voltage correction loop, and the output of the correction link is the correction voltage Uxiuzheng , which is added to the droop control voltage loop, the source side optimizes the droop control through the bidirectional DC-DC converter, performs reasonable power distribution, and bears all the load-side demand current.

[0085] Step 4: During the battery balancing process, dynamically obtain the battery status and get the SOC i , SOC avg , SOC max , SOC min , ΔSOC max , ΔSOC i After a period of time, the source-load separation, balancing mode judgment, and load-side balancing current I are re-performed according to the above parameters. equ_i Calculation and source side droop coefficient Droop i Update, that is, adjust the energy balance in real time until the balance is completed.

[0086] In order to show the significant effect of the present invention, this embodiment gives some experimental results obtained by using the embodiment. Under the condition that the internal balanced bus voltage is given as 100V, for four 48V battery packs, first, the current SOC state of each battery pack is manually set, and the change of the SOC state is set on this basis, and the balanced parameters and balanced state of each battery pack are observed. The setting change of the SOC state is as follows: Fig.10 The automatic change of the equalization parameters according to the SOC change obtained by the present invention is shown in FIG. Fig.11 The experimental results are shown in Fig.12 shown. Fig.12 are the low-voltage side currents of battery groups 1, 2, and 3 and the DC bus voltage. When the balancing threshold is met, each battery group is in the corresponding balancing mode. When the SOC of each battery group changes, the source-load situation changes accordingly, and the energy exchange situation changes from 1 (source) to 3 (load) to 2 (source) to 2 (load), and then to 3 (source) to 1 (load). In addition, according to the different droop coefficients on the source side, the power borne by each phase is also different. For example, in the 2-to-2 stage, the output currents of battery group 1 and battery group 2 as sources are 3A and 2.2A respectively. The balancing current on the load side also changes dynamically. For example, in the 3-to-1 case, the balancing current of battery group 4 as a load is 2.59A, and in the 2-to-2 case, the balancing current of battery group 4 as a load is 2.09A. At the same time, after optimizing the droop control, the internal balanced DC bus voltage has been maintained at a given voltage of about 100V. By Fig.11 , Fig.12 It can be seen that the source-load situation, balancing parameters and balancing state can be automatically judged according to the SOC change, and can be quickly and dynamically adjusted to keep the balance between battery packs in a reasonable state and maintain a stable energy balance state.

[0087] The experimental results obtained in this embodiment can illustrate that when the SOC difference between battery groups reaches a set threshold, the method described in the present invention can automatically determine parameters such as the balancing mode and balancing current, and dynamically control the energy balancing speed between any battery groups, greatly improving the flexibility, balancing speed and efficiency of SOC balancing between battery groups.

[0088] The present invention can be implemented in other specific forms without departing from its spirit or essential characteristics. The described embodiments are considered to be illustrative and non-restrictive in all aspects, for example:

[0089] 1) The non-isolated bidirectional DC-DC topology used in the inter-group balancing topology is not limited to the configuration in the embodiment;

[0090] 2) The parameters of the battery pack and other relevant charging and discharging data are not limited to those used in the embodiments;

[0091] 3) The multiple thresholds in the equalization mode judgment are not limited to those used in the embodiment;

[0092] 4) The domain, fuzzy subset, fuzzy rule table, etc. in fuzzy control are not limited to those used in the embodiments;

[0093] 5) Balancing parameters, such as balancing current, droop coefficient, etc. are not limited to those used in the embodiments. Therefore, the scope of the present invention is indicated by the appended claims rather than the above description. All changes within the meaning and scope of the equivalent technical solutions of the claims are included in their scope.

Claims

1. An active energy balancing method for source-load separation between battery groups, characterized in that: The active energy balancing method performs active energy balancing between battery packs based on connecting battery packs using a DC-DC bidirectional converter balancing topology hardware platform; The active energy balancing method uses fuzzy control to obtain the load-side (energy absorption side) balancing current and the source-side (energy output side) droop coefficient after performing source-load separation and balancing mode judgment according to the characteristics of the balancing topology, and then combines the optimized droop control to perform energy balancing, including the following steps: (1) Estimate the power of each phase battery group and use the source-load separation method to divide each phase battery group into the source side (energy output side) and the load side (energy absorption side). The corresponding converters are the source side converter and the load side converter respectively; (2) Adopting a balancing strategy to determine whether balancing is enabled for each phase of the battery pack; (3) Fuzzy control combined with optimized droop control method is used. After balancing is turned on, fuzzy control is used to calculate the balanced current I on each load side. equ_i Then determine the total load current on the bus, and use fuzzy control to calculate the droop coefficient of each source side. i , combined with optimized droop control to distribute power to each source side, so that the source side transmits energy to the load side for energy balance; the fuzzy control combined with optimized droop control method is: after turning on the balance, on the load side, the SOC i and ΔSOC i As the input of the load-side fuzzy controller, the load-side balanced current I is obtained according to the load-side fuzzy rules. equ_i ; On the source side, SOC i and ΔSOC i As the input of the source side fuzzy controller, the source side droop coefficient Droop is obtained according to the formulated source side fuzzy rule. i , add a balanced internal bus voltage correction link in the source side droop control link, the input of this link is the difference between the bus voltage set value and the actual value, adopts proportional integral control, the output is the bus voltage correction value, and the bus voltage set value can be changed; (4) During the battery balancing process, the balancing mode and load-side balancing current I are dynamically updated according to the battery status. equ_i And the source side droop coefficient Droop i .

2. The active energy balancing method for source-load separation between battery packs as claimed in claim 1, characterized in that: The DC-DC bidirectional converter balancing topology is a high-voltage side common DC bus structure, any number of battery packs are connected to the same number of low-voltage sides of each converter, and the high-voltage sides of each converter are connected in parallel to establish an internal balancing bus with no other loads on it.

3. The active energy balancing method for source-load separation between battery packs as claimed in claim 1, characterized in that: The source-load separation method is: using the SOC of each phase battery pack i , calculate the average SOC of all battery packs avg , Maximum SOC max , minimum SOC min , range ΔSOC max , SOC of each phase battery pack i With SOC avg The difference ΔSOC i , according to ΔSOC i The positive or negative value of ΔSOC determines whether each phase battery pack is controlled as the source side or the load side. i is a positive value, the battery pack of this phase is controlled as the source side; ΔSOC i If it is a negative value, the battery pack of this phase is controlled as the load side.

4. The active energy balancing method for source-load separation between battery packs as claimed in claim 1, characterized in that: The balancing strategy is: i Set different thresholds for opening and closing balancing. When the threshold conditions are met, the balancing of the battery pack of this phase is turned on or off. max Set the threshold for turning on and off balancing. If the threshold condition is met, balancing is turned on or off for the battery group with the largest and smallest power.

5. The active energy balancing method for source-load separation between battery packs as claimed in claim 1, characterized in that: During the balancing process, the battery pack status is updated in real time. The balancing mode and load-side balancing current I are re-determined based on the battery status every once in a while. equ_i Calculation and source side droop coefficient Droop i Updates.

Images